feat: outils IA de lecture et d'edition des classeurs existants #153

A6 : list_xlsx_sheets (noms + dimensions + truncated), xlsx_to_markdown (table bornee 100x20 pour le contexte LLM), update_xlsx_cells et append_xlsx_rows (WRITE + confirmation, via edit_xlsx_cells : verrou, ecriture atomique, neutralisation des formules, 409 lossy). Les lignes ajoutees sont coercees comme dans le viewer (A10). Labels de step ai.step.xlsx_* FR/EN ; update_xlsx_cells/append_xlsx_rows branches sur MUTATING_TOOLS et FILE_WRITE_TOOLS (refresh viewer via obsigate:file-written). Tests : test_spreadsheet_tools.py 17 (risque, confirmation, persistance, garde formule heritee, coercion) + contre-preuve (cells vide -> 2 echecs). ruff 0, mypy 0, i18n parity, validate-imports. 🤖 Generated with Codebuff Co-Authored-By: Codebuff <[email protected]>
feat: chargement à la demande des lignes cachées des tableurs #153
2026-09-28 12:05:16 -04:00 · 2026-09-28 11:52:42 -04:00 · 2026-09-28 10:38:24 -04:00 · 2026-09-27 22:43:24 -04:00 · 2026-09-27 20:39:12 -04:00 · 2026-09-27 14:46:09 -04:00
138 changed files with 13914 additions and 4707 deletions
@@ -12,8 +12,12 @@ OBSIGATE_ADMIN_PASSWORD=chab30
 # (ex. 0.0.0.0) sauf si l'on force l'opt-in ci-dessous. À réserver au local.
 # OBSIGATE_ALLOW_INSECURE=false

-# Sécurité des cookies (activer si derrière HTTPS)
-# OBSIGATE_SECURE_COOKIES=false
+# Sécurité des cookies : true|false|auto (défaut : auto — Secure si la
+# requête arrive en https, sinon pas de flag ; les navigateurs ignorent les
+# cookies `Secure` en HTTP, ce qui casserait les logins en local).
+# Derrière un reverse proxy qui termine TLS, auto suffit avec
+# OBSIGATE_TRUST_PROXY=true (X-Forwarded-Proto honoré).
+# OBSIGATE_SECURE_COOKIES=auto

 # Tokens TTL en secondes
 # OBSIGATE_ACCESS_TOKEN_TTL=31536000000  # 1000 ans
@@ -23,6 +27,8 @@ OBSIGATE_ADMIN_PASSWORD=chab30
 # OBSIGATE_LOGIN_MAX_ATTEMPTS=10
 # OBSIGATE_ACCOUNT_MAX_ATTEMPTS=10
 # OBSIGATE_LOGIN_WINDOW_SECONDS=900
+# Compteurs partagés/persistants (SQLite WAL, multi-workers) — défaut : mémoire.
+# OBSIGATE_RATELIMIT_DB=data/ratelimit.db

 # IP client derrière un reverse proxy (fait confiance à X-Forwarded-For)
 # OBSIGATE_TRUST_PROXY=false
@@ -43,8 +43,12 @@ jobs:
          node tests/frontend/forge-completion.test.mjs
          node tests/frontend/config-mobile.test.mjs
          node tests/frontend/settings-order-avatar.test.mjs
+          node tests/frontend/mobile-toolbar.test.mjs
+          node tests/frontend/pretty.test.mjs
+          node tests/frontend/media-viewer.test.mjs
+          node tests/frontend/mfa-settings.test.mjs

-      - name: Frontend JSDOM tests (PaneManager + Excalidraw + Plugins + AI + SW + Collab + Mobile + Semantic + Desktop + Inline edition)
+      - name: Frontend JSDOM tests (PaneManager + Excalidraw + Plugins + AI + SW + Collab + Mobile + Semantic + Desktop + Inline edition + Upload + XLSX)
        run: |
          cd tests/frontend
          if [ -d node_modules ]; then
@@ -62,6 +66,9 @@ jobs:
            node toolbar-order.test.mjs
            node editor-inline.test.mjs
            node ai-quick-actions.test.mjs
+            node upload.test.mjs
+            node config-ai-keys.test.mjs
+            node xlsx-viewer.test.mjs
          else
            echo "tests/frontend/node_modules missing - installing jsdom"
            npm install --no-audit --no-fund --silent
@@ -79,6 +86,9 @@ jobs:
            node toolbar-order.test.mjs
            node editor-inline.test.mjs
            node ai-quick-actions.test.mjs
+            node upload.test.mjs
+            node config-ai-keys.test.mjs
+            node xlsx-viewer.test.mjs
          fi

  # ── Tests ─────────────────────────────────────────────────────────
@@ -122,14 +132,30 @@ jobs:

      - name: Install dependencies
        run: |
-          pip install bandit pip-audit
+          pip install bandit pip-audit semgrep
          pip install -r backend/requirements.txt

-      - name: Bandit (SAST)
-        run: bandit -r backend/ --skip B101,B110,B310 || echo "bandit found issues (non-blocking)"
+      - name: Bandit (SAST, bloquant — #87)
+        # B105 est exclu (aligné avec [tool.bandit] de pyproject.toml :
+        # faux positifs systématiques sur les noms de variables) ; les rares
+        # vrais positifs restants portent un `# nosec` justifié inline.
+        run: bandit -r backend/ --skip B101,B105,B110,B310

-      - name: Pip-audit (dependency vulnerabilities)
-        run: pip-audit || echo "pip-audit found vulnerabilities (non-blocking)"
+      - name: Semgrep (SAST local, bloquant — #87)
+        # Règles 100 % locales (semgrep-rules/, 8 règles) : aucun
+        # téléchargement de registre (runner au réseau fragile).
+        run: semgrep --config semgrep-rules/ backend/
+
+      - name: Pip-audit (bloquant — #87)
+        # Bloquant depuis T6 (#87) : dépendances qualifiées (mistune 3.3.3,
+        # python-multipart 0.0.31, weasyprint 70, mcp 1.28.1, fastapi 0.141.1
+        # + starlette 1.7.0, setuptools 84 — suite complète verte + 0 vuln).
+        # Seule exception documentée : PYSEC-2026-1325 (ecdsa, Minerva) —
+        # aucun correctif upstream ET ObsiGate ne signe/vérifie qu'en HS256
+        # (backend/auth/jwt_handler.py), les chemins ECDSA P-256 ne
+        # s'exécutent jamais.
+        # NOTE runner Gitea Act (BUG-083) : aucun `#` dans le `run:`.
+        run: pip-audit --ignore-vuln PYSEC-2026-1325

  # ── Docker build ──────────────────────────────────────────────────
  build:
@@ -191,6 +217,9 @@ jobs:
          npm ci
          npx playwright install --with-deps chromium

+      - name: Npm audit (bloquant — #87, 0 dépendance prod hors Playwright)
+        run: npm audit --omit=dev
+
      - name: Start ObsiGate
        run: |
          docker rm -f obsigate-e2e 2>/dev/null || true
@@ -31,6 +31,20 @@ desktop/backend/
 desktop/frontend/
 backend/VERSION

+# Artefacts générés par les runs E2E (excalidraw crée ces diagrammes)
+test_vault/IT/e2e-diagram-*.excalidraw
+
+# Fixtures de test locales non versionnées (~200 Mo, pas de fixture CI).
+# Aucun test/CI ne les référence : les tests unitaires génèrent leurs fixtures
+# dans tmp_path (tests/conftest.py), et l'E2E n'utilise que les fixtures
+# committées (test_vault/sample-*.{mp3,png,svg,webm,pdf}, test_dir/*.md).
+# → à committer volontairement : `git add -f <chemin>`.
+test_dir/music/
+test_dir/video/
+test_vault/images/
+test_vault/markdown/
+test_vault/budget.xlsx
+
 # Tauri updater signing keys (private key — never commit)
 desktop/*.key
 desktop/*.key.pub
@@ -6,7 +6,7 @@ Format basé sur [Keep a Changelog](https://keepachangelog.com/fr/1.1.0/),
 et [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

 > **En cours de développement** : les changements à venir sont listés dans la section
-> [Unreleased](#unreleased). La dernière version livrée est **2.22.0**.
+> [Unreleased](#unreleased). La dernière version livrée est **2.33.0**.

 ---

@@ -14,6 +14,702 @@ et [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

 ---

+## [2.33.0] — 2026-09-28
+
+---
+
+## [2.32.0] — 2026-09-28
+
+---
+
+## [2.31.0] — 2026-09-28
+
+### Correction
+
+- **BUG-090 — troncature silencieuse d'une feuille `.xlsx` au-delà de
+  500 lignes × 40 colonnes.** `render_sheets()` renvoie les dimensions
+  déclarées par la feuille (`total_rows`/`total_cols`), les plafonds du
+  moteur (`max_rows`/`max_cols`) et un flag `truncated` : la visionneuse
+  affiche un bandeau « Feuille tronquée — 500 lignes affichées sur 520 »
+  (i18n FR/EN) au lieu de présenter une table courte comme complète. La
+  ligne d'en-têtes est désormais figée au défilement vertical (`thead`
+  sticky, `top: auto` sur les numéros de ligne pour éviter leur
+  empilement en haut à gauche). *#153 A8/R5.*
+
+### Ajouté
+
+- **#153 A9 — chargement paresseux d'une feuille par fenêtres.**
+  `GET /api/file/{vault}/xlsx/sheet?sheet=&offset=&limit=` renvoie un
+  bloc de lignes (`XlsxSheetWindowResponse`, plafond 1 000 lignes par
+  requête, `has_more` de pagination) avec les **vraies** coordonnées A1
+  et numéros de ligne de la feuille — une fenêtre se comporte exactement
+  comme le rendu complet. Erreurs typées : 404 feuille inconnue, 415
+  fichier non-`.xlsx`. La lecture des valeurs calculées en cache (#153
+  A12) s'applique aussi aux fenêtres.
+- **#153 A9bis — « Charger la suite » sous une feuille tronquée.** Un
+  pied de page annonce la progression et fetch la fenêtre suivante au
+  clic ou à l'approche du bas du tableau (sentinelle de défilement).
+  Les lignes ajoutées passent par le même pipeline d'édition que le
+  rendu initial : éditables et sauvegardables immédiatement. Un fetch
+  échoué restore le bouton (retry possible) ; feuille complète → pied
+  de page masqué.
+
+---
+
+## [2.30.0] — 2026-09-27
+
+### Correction
+
+- **BUG-089 — un reindex manuel ne reconstruisait pas l'index inversé.**
+  `reload_index()` / `reload_single_vault()` remplacent l'entrée de vault
+  en bloc, ce qui n'émet pas les notifications incrémentales : la
+  recherche TF-IDF continuait de servir un index périmé après un
+  reindex. Les deux fonctions appellent désormais `init_inverted_index()`.
+  Au passage, `backend/search.py` lisait l'index via
+  `from backend.indexer import index` — une liaison **par valeur** du
+  dict : un rechargement du module `backend.indexer` recréait le dict
+  côté indexer alors que la recherche écrivait dans l'ancien, et
+  l'index inversé n'indexait plus rien. Tous les accès passent par
+  `_indexer.index`. *Trouvé en écrivant le test de recherche d'A5 : il
+  passait isolément et échouait en suite complète selon l'ordre.*
+
+### Ajouté
+
+- **#153 A5 — les tableurs sont indexés par leur contenu.**
+  `extract_indexable_text()` extrait les noms de feuilles et les 20
+  premières lignes (plafond 5 000 caractères, 20 feuilles) pour le
+  TF-IDF et la recherche sémantique. Un mot tapé dans une cellule rend
+  désormais le classeur trouvable ; la lecture binaire pour l'affichage
+  est inchangée et un classeur chiffré/corrompu s'indexe par son seul
+  nom.
+- **#153 A10 — la saisie est typée comme dans Excel.** Une valeur
+  `TRUE`/`FAUX`/`OUI`/`NON` devient un booléen, une date `JJ/MM/AAAA`
+  (avec `HH:MM` optionnel) devient une vraie date — et dans l'ordre
+  français : `01/02/2026` est le 1ᵉʳ février. Une saisie ressemblant à
+  une formule n'est jamais convertie.
+- **#153 A12 — la valeur calculée s'affiche sous la formule.** Quand une
+  cellule porte encore le résultat de son dernier calcul Excel, celui-ci
+  s'affiche dans une ligne discrète sous la formule. La seconde lecture
+  `data_only=True` n'a lieu que si l'archive contient réellement une
+  valeur en cache, et toute erreur retombe sur l'affichage formules seul.
+  Info-bulle traduite FR/EN (`xlsx.cached_value_title`).
+
+---
+
+## [2.29.0] — 2026-09-27
+
+### Correction
+
+- **BUG-084 — l'index inversé conservait des documents fantômes après la
+  suppression d'une vault.** `remove_vault_from_index()`
+  (`backend/indexer.py`) ne notifiait pas le hook incrémental : après
+  suppression d'une vault, ses documents restaient dans l'index inversé
+  (`postings`, `doc_info`, `doc_vault`, `vault_docs`) et continuaient de
+  correspondre aux recherches pour une vault inexistante — seul un reindex
+  manuel les effaçait. Le correctif déclenche
+  `_on_index_change('remove', …)` pour chaque fichier de la vault, et
+  `_remove_doc_internals()` supprime désormais la clé `vault_docs` dont le set
+  devient vide (c'est un `defaultdict` : une lecture la recréait).
+  Test : `TestVaultRemovalPurgesInvertedIndex` (contre-preuve : échoue sans le
+  correctif).
+
+### Maintenance
+
+- **Index inversé — `is_stale()` renommé `is_ready()`.** La relecture de
+  `plan.md` a établi que les étapes 6 et 7 (suppression du cooldown et du hack
+  de coalescence) étaient **déjà livrées** : ni `_last_rebuild`, ni
+  `_rebuild_cooldown`, ni `_source_generation`, ni `_on_vault_change` ne
+  subsistent. `is_stale()` ne mesurait donc plus aucune staleness — il
+  indiquait seulement si l'index initial était construit, sous un nom
+  trompeur. Renommé `is_ready()`, cohérent avec le `is_ready()` de
+  `SemanticIndex` ; l'alias `is_stale()` de `SemanticIndex`, sans appelant, est
+  supprimé. `/api/diagnostics` expose désormais `is_ready` (libellé « Index
+  prêt » côté `frontend/js/config.js`). Tests :
+  `test_is_ready_tracks_initial_build`, `test_is_ready_survives_incremental_updates`.
+
+- **`plan.md` recalibré.** Le fichier est désormais marqué « livré » et
+  suivi d'une section « État réel » : le code a divergé du plan sur quatre
+  points (pas de repli `_needs_rebuild`, `_ready` au lieu de `doc_count == 0`,
+  `rebuild()` conservé au démarrage, `is_stale()` repurposé). Les extraits de
+  code du plan sont explicitement signalés comme ne décrivant pas le code
+  actuel.
+
+- **Fixtures de test locales exclues du suivi Git.** `test_dir/music/`,
+  `test_dir/video/`, `test_vault/images/`, `test_vault/markdown/` et
+  `test_vault/budget.xlsx` (~200 Mo) sont ajoutés au `.gitignore` : aucun test
+  ni job CI ne les référence — les tests unitaires génèrent leurs fixtures dans
+  `tmp_path` et l'E2E n'utilise que les fixtures committées
+  (`test_vault/sample-*.{mp3,png,svg,webm,pdf}`, `test_dir/*.md`). Ils
+  restaient non suivis et polluaient `git status`.
+
+### Sécurité
+
+- **BUG-088 — plus d'injection de formule via la visionneuse Excel.** Une
+  saisie `=cmd|'/c calc'!A1` (ou `@…`) était stockée comme **formule** par
+  openpyxl, donc exécutée par Excel à la réouverture du fichier (DDE).
+  `edit_xlsx_cells` force maintenant le type texte (`cell.data_type = "s"`)
+  pour toute valeur commençant par `=` ou `@` ; l'API accepte
+  `allow_formula: true` et la visionneuse expose un bouton `f(x)`
+  (opt-in, état de session, jamais persisté). `+`/`-` restent des nombres.
+- **BUG-087 — écriture concurrente d'un classeur.** `load_workbook()` →
+  `save()` n'était pas sérialisé : deux sauvegardes simultanées (deux
+  onglets, l'agent IA et la visionneuse) faisaient gagner la dernière, en
+  silence. Verrou par chemin (`backend/services/mutations.py::_xlsx_write_lock`,
+  timeout 15 s) autour du cycle lecture → édition → remplacement ; attente
+  dépassée → **409** `conflict`. L'endpoint `PUT …/xlsx/save` est devenu
+  synchrone pour que l'attente s'exécute dans le threadpool.
+
+### Corrigé
+
+- **BUG-085 — la perte de données à l'enregistrement d'un `.xlsx` est
+  annoncée, plus silencieuse.** `GET /api/file/{vault}` renvoie
+  `xlsx_lossy_features` (éléments qu'un round-trip openpyxl perd) ; la
+  visionneuse affiche un bandeau listant ces éléments et la première
+  sauvegarde demande confirmation avant de renvoyer `force: true`. Sans
+  `force`, l'API répond **409** `xlsx_lossy_content` avec
+  `details.features`. Périmètre **remesuré** sur openpyxl 3.1.5 : graphiques,
+  images, dessins et tableaux croisés sont bien préservés ; sont perdus les
+  valeurs calculées en cache, slicers/chronologies, contrôles de formulaire,
+  connexions/requêtes, custom XML, signature numérique, commentaires
+  enrichis et macros.
+- **BUG-086 — écriture atomique des classeurs.** `wb.save()` écrivait en
+  place sur le fichier du vault : un plantage laissait un `.xlsx` tronqué.
+  L'écriture passe désormais par un `.tmp` puis `os.replace()` (le backup
+  `.bak` est inchangé, le `.tmp` est ignoré par le watcher).
+- Le handler global `ServiceError` expose maintenant `code` et `details` dans
+  la réponse JSON, et `api()` (frontend) les propage sur l'Error — nécessaire
+  pour que le client distingue un 409 de confirmation d'une autre erreur.
+
+### Ajouté
+
+- **#153 (P0) — tests de la visionneuse Excel.**
+  `tests/frontend/xlsx-viewer.test.mjs` (10 tests JSDOM : bannière,
+  confirmation + reprise `force`, refus, toggle `f(x)`, payload de
+  sauvegarde) et `tests/e2e/xlsx-viewer.spec.js` (3 tests Playwright sur la
+  fixture `test_vault/sample-xlsx-lossy.xlsx`) ; la suite JSDOM est branchée
+  dans le CI.
+
+### Documentation
+
+- **#153 — Visionneuse & édition XLSX : audit complet et backlog de
+  complétude.** La visionneuse `.xlsx` livrée par #152 a été auditée couche
+  par couche (lecture `backend/xlsx_reader.py`, écriture
+  `backend/services.mutations.edit_xlsx_cells`, UI `renderXlsxViewer`,
+  indexation, outils IA, tests). Bilan : la grille de valeurs est éditée
+  correctement (sécurité, backup, audit, échappement HTML), mais l'ensemble
+  supporté est étroit, une partie du classeur est perdue à l'enregistrement,
+  les tableurs sont **invisibles pour la recherche** et l'IA ne sait que les
+  **créer**. Ouverture de l'item **#153** dans `docs/ROADMAP.md` (17
+  sous-tâches suivies **A1 → A17** ; **P0 livré**, reste P1 recherche/IA/UX
+  puis P2 étendu) et création de la fiche
+  [docs/features/xlsx-viewer.md](docs/features/xlsx-viewer.md) : cartographie
+  du code, limites par couche, tableau des risques R1-R5 et critères
+  d'acceptation par sous-tâche.
+
+---
+
+## [2.28.16] — 2026-09-27
+
+---
+
+## [2.28.15] — 2026-09-27
+
+### Sécurité
+
+- **#87 T6 — dépendances qualifiées, `pip-audit` bloquant (0 vulnérabilité).**
+  mistune 3.0.2 → 3.3.3 (XSS/ReDoS/DoS dans le moteur de rendu),
+  python-multipart 0.0.9 → 0.0.31, weasyprint 69 → 70, mcp 1.9.4 → 1.28.1,
+  fastapi 0.110.3 → 0.141.1 + starlette 0.37.2 → 1.7.0, setuptools 84 ;
+  `cast(str, …)` aux 3 sites d'appel mistune (typage 3.3 resserré). Suite
+  complète 1359 passed, ruff/mypy 0. Seule exception : PYSEC-2026-1325
+  (ecdsa, Minerva) — aucun correctif upstream ET JWT exclusivement HS256
+  (`backend/auth/jwt_handler.py`), les chemins ECDSA P-256 ne s'exécutent
+  jamais → `--ignore-vuln` documenté.
+
+- **#87 T7 — semgrep SAST local bloquant (8 règles, 0 finding).**
+  Ruleset `semgrep-rules/` (eval/exec, shell=True, os.system, pickle,
+  yaml.load sans Loader, verify=False, Markup, mktemp) — 100 % local,
+  aucun registre réseau (runner au réseau fragile). Trivy écarté :
+  binaire + base de vulnérabilités à télécharger à chaque run, couche
+  Python déjà couverte par `pip-audit` bloquant (image = slim + 4 libs).
+
+- **#87 T8 — fin BUG-034 : cookies Secure auto, CORS same-origin explicite.**
+  `OBSIGATE_SECURE_COOKIES=true|false|auto` (défaut auto : Secure en https,
+  sinon rien — logins http locaux préservés ; `X-Forwarded-Proto` honoré
+  sous `TRUST_PROXY`, avertissement démarrage affiné, `TRUST_PROXY=true`
+  dans le compose prod) ; `CORSMiddleware` same-origin explicite (sûr :
+  web et desktop Tauri same-origin, API directe hors navigateur) ;
+  `style-src 'unsafe-inline'` conservé et assumé (189 attributs `style=` +
+  343 `el.style` — suppression = réécriture complète, risque nul côté
+  exécution une fois `script-src` verrouillé en T5c).
+
+---
+
+## [2.28.14] — 2026-09-27
+
+---
+
+## [2.28.13] — 2026-09-27
+
+### Modifié
+
+- **#87 (T5c) — `script-src` sans `'unsafe-inline'`.**
+  Seuls les scripts avec nonce frais (`backend/csp.py`, T5b) ou servis par
+  `'self'`/CDN listés s'exécutent ; `style-src` garde `'unsafe-inline'`
+  (chantier séparé). Vérifié : `test_csp_nonce.py` 5/5, 0 handler inline
+  restant dans les pages HTML (propriétés `onXxx = fn` en JS non concernées
+  par la CSP).
+
+---
+
+## [2.28.12] — 2026-09-27
+
+### Corrigé
+
+- **BUG-081 — `GET /api/auth/mfa/status` → 500 quand l'auth est désactivée.**
+  Le pseudo-user `anonymous` (auth désactivée, mode E2E/CI) n'a aucune entrée
+  en store : `get_user(...)` → `None` puis `AttributeError` sur `user.get`.
+  Garde `None` → payload « MFA désactivé » (`mfa_enabled: false`,
+  `totp_enabled: false`, `webauthn_credentials: 0`). Test : `tests/test_mfa.py`
+  (`TestMfaStatusAuthDisabled`, échoue en 500 sans le correctif).
+
+---
+
+## [2.28.11] — 2026-09-27
+
+---
+
+## [2.28.10] — 2026-09-27
+
+### Corrigé
+
+- **BUG-083 — job CI `security` rouge : le runner tronquait le `#` du `run:` pip-audit.**
+  Le runner Gitea Act coupe naïvement au premier `#` (même entre
+  guillemets) : `echo "... see #87)"` devenait une citation non fermée
+  (`unexpected EOF while looking for matching '"'"`). Seul `run:` du
+  workflow avec un `#` ; l'echo n'a plus de `#` (réf `#87` en commentaire
+  YAML, jamais vu par le shell). Garde-fou : `tests/test_ci_workflow.py`
+  (aucun `#` dans le code des `run:`, `upload.test.mjs` verrouillé dans
+  l'étape JSDOM — BUG-082).
+
+---
+
+## [2.28.9] — 2026-09-27
+
+### Corrigé
+
+- **BUG-082 — CI `lint` rouge : suites frontend exigeant `jsdom`.**
+  `tests/frontend/upload.test.mjs` puis `config-ai-keys.test.mjs` (imports
+  statiques `jsdom`, introduits par `#89`) étaient exécutés dans l'étape
+  frontend racine où `jsdom` n'est jamais installé (`ERR_MODULE_NOT_FOUND`,
+  rouge depuis `7bee4a2`). Déplacés dans l'étape JSDOM (les deux branches,
+  après install si besoin) ; garde-fou `tests/test_ci_workflow.py` :
+  aucun fichier de l'étape racine ne doit importer `jsdom` statiquement.
+
+---
+
+## [2.28.8] — 2026-09-27
+
+### Corrigé
+
+- **BUG-080 — harnais E2E local anti-blocage (plus de run pendu toute la nuit).**
+  `run-e2e-local.ps1` : Playwright lancé via `node` direct sur la CLI locale
+  (jamais de prompt interactif, `Start-Process` ne sachant pas exécuter `npx` ;
+  paramètre `$Arguments`, `$Args` étant une variable automatique qui l'écraserait),
+  installation Chromium sautée si déjà présent (`E2E_INSTALL_BROWSERS=1`
+  pour forcer), étapes `install`/`test` bornées (`E2E_TIMEOUT_SEC`,
+  défaut 1800 s / 600 s, exit 124 au dépassement — au-delà du globalTimeout
+  pour un abandon propre avec rapport) ; `run-e2e-local.sh` : `npx --yes` +
+  mêmes bornes ; `playwright.config.ts` : `globalTimeout` (25 min en local,
+  30 min en CI, `E2E_GLOBAL_TIMEOUT_MS` pour surcharger) ; `e2e-server.ps1` :
+  pidfile resynchronisé sur le vrai propriétaire du port et `stop` qui tue
+  l'arbre complet (fini les serveurs orphelins qui squattent le port 2029).
+  Garde-fous : `tests/test_e2e_harness.py` (8 tests).
+
+---
+
+## [2.28.7] — 2026-09-26
+
+### Ajouté
+
+- **#87 (T5b) — nonces CSP prêts pour la bascule (sans changement).**
+  Nonce frais par réponse dans `script-src` (`backend/csp.py`), injecté
+  dans les 6 pages HTML servies (dont la nouvelle route
+  `/excalidraw-editor.html`, utilisée par l'iframe du viewer au lieu de
+  `/static/`) et la page de partage ; `tests/test_csp_nonce.py` (unicité,
+  concordance en-tête/HTML). `unsafe-inline` conservé jusqu'en T5c.
+
+---
+
+## [2.28.6] — 2026-09-26
+
+### Modifié
+
+- **#87 (T5a) — 16 handlers inline convertis en listeners (CSP inchangée).**
+  `onclick`/`onerror` de `index.html` et des vues JS (`config`, `plugins`,
+  `sync`, `viewer`, `auth`) remplacés par `addEventListener` ; suites
+  frontend vertes (unit, ai, config-mobile, pdf-viewer, mfa-settings,
+  sidebar-filters).
+
+---
+
+## [2.28.5] — 2026-09-26
+
+---
+
+## [2.28.4] — 2026-09-26
+
+### Ajouté
+
+- **#87 (T4) — E2E XSS et serveur E2E piloté.**
+  `tests/e2e/xss.spec.js` : page publique `/s/{token}` (titre/frontmatter
+  hostile échappé, JSON neutralisé, aucun JS exécuté) et lecteur markdown
+  (sanitizer, aucun `on*`/`javascript:` vivant) — 2/2 verts en local.
+  `scripts/e2e-server.ps1` : `start|stop|status|logs` avec progression
+  visible et fichier PID (fini les serveurs orphelins sur le port 2029).
+
+---
+
+## [2.28.3] — 2026-09-26
+
+### Ajouté
+
+- **#87 (T3) — cookies `Secure` et CORS explicites.**
+  Helper `is_secure_cookies()` centralisé (défaut `false` conservé pour ne
+  pas casser les logins HTTP locaux) + avertissement au démarrage sur bind
+  non-loopback sans `Secure` ; `tests/test_security_headers.py` atteste
+  l'absence de CORS permissif (same-origin par défaut du navigateur) et les
+  en-têtes de durcissement.
+
+---
+
+## [2.28.2] — 2026-09-26
+
+### Ajouté
+
+- **#87 (T2) — tests de durcissement : concurrence et regex.**
+  `tests/test_hardening_concurrency.py` : créations/mises à jour/`login
+  failures` concurrents sur `users.json` (zéro mise à jour perdue, JSON
+  toujours récupérable) et budget temps de la politique ReDoS (motifs
+  catastrophiques rejetés en < 1 s, motifs acceptés < 5 s sur 200 Ko).
+
+---
+
+## [2.28.1] — 2026-09-26
+
+### Modifié
+
+- **#87 (T1) — CI sécurité durcie.**
+  `bandit` passe en bloquant (`# nosec` justifiés : SHA1 non-crypto,
+  subprocess git à argv fixe, `saxutils.escape` sans parsing ; B105 exclu
+  comme `pyproject.toml`) ; `npm audit --omit=dev` bloquant (0 faille) ;
+  les 5 suites frontend hors CI (`upload`, `pretty`, `media-viewer`,
+  `mfa-settings`, `config-ai-keys`, vertes en local) rejoignent le job
+  `lint`. `pip-audit` reste consultatif (upgrades starlette/weasyprint à
+  qualifier, chantier dédié).
+
+---
+
+## [2.28.0] — 2026-09-26
+
+### Modifié
+
+- **#85 (T10) — persistance d'état et clôture de la refonte architecturale.**
+  Verrous `RLock` sur les stores JSON sans protection (`revoked_tokens`,
+  `shares`, `webhooks` + secrets, clés d'outils) avec tests de concurrence
+  (`tests/test_store_locks.py` — pertes prouvées sans verrou) ; rate-limit
+  auth persisté en option (`OBSIGATE_RATELIMIT_DB`, SQLite WAL, sémantique
+  identique, défaut mémoire inchangé, `tests/test_ratelimit_store.py`).
+  Contrat `tools/registry.py` audité (permissions/quotas/redaction déjà
+  câblés, rien à coder). Index non persisté : rebuild différentiel #86
+  suffisant (décision documentée). Fiche `docs/features/archi-refonte-85.md`,
+  #85 sorti du backlog (index roadmap).
+
+## [2.27.12] — 2026-09-26
+
+### Modifié
+
+- **#85 (T9) — extraction realtime + render hors du monolithe `backend/main.py`.**
+  Le stream SSE `/api/events` et le WebSocket `/ws/collab/*` sont servis par
+  `backend/routers/realtime.py`, le pipeline markdown (mistune, wikilinks,
+  slugs, sanitizer) par `backend/render.py` (imports directs, plus de
+  couplage différé). `main.py` (4 827 → ~760 lignes) ne contient plus que
+  l'assemblage : lifespan, middlewares, montage des 16 routers, racine
+  `/api`, statique/SPA et cales de compatibilité testées.
+
+## [2.27.11] — 2026-09-26
+
+### Modifié
+
+- **#85 (T8) — extraction vaults/history/conflicts hors du monolithe `backend/main.py`.**
+  13 routes servies par `backend/routers/vaults.py`, `history.py` et
+  `conflicts.py` ; `VaultInfo`/`BookmarkToggleRequest` dans `schemas.py`,
+  handle watcher partagé dans `backend/watcher_state.py`.
+  `tests/test_api_main.py` importe `humanize_mtime` depuis son module
+  canonique (`services.recent`).
+
+## [2.27.10] — 2026-09-26
+
+### Modifié
+
+- **#85 (T7) — extraction du domaine `config` hors du monolithe `backend/main.py`.**
+  `/api/config`, ai-keys (get/post/delete/test), tool-keys (×3), ai-models,
+  diagnostics et dashboard sont servis par `backend/routers/config.py`
+  (`_FALLBACK_MODELS`, store clés et config déplacés ; `main` réimporte
+  `_load_config` pour son lifespan, les fixtures de tests inchangées).
+  `tests/test_ai_models.py` patch désormais la référence du router.
+
+## [2.27.9] — 2026-09-26
+
+### Modifié
+
+- **#85 (T6c) — extraction media/pdf/export/guide hors du monolithe `backend/main.py`.**
+  file/pdf, exports (html/md-bundle/epub), guide/download, pdf/stream|info,
+  image, media+thumb, attachments (rescan/stats), vault settings (get/post/all)
+  et vault files sont servis par `backend/routers/files_media.py` ; le helper
+  Range partagé vit dans `backend/routers/helpers.py` (tags OpenAPI inchangés,
+  tests statiques frontend `media-viewer`/`image-viewer` réalignés).
+
+## [2.27.8] — 2026-09-26
+
+### Modifié
+
+- **#85 (T6b) — extraction mutations fichiers/dossiers hors du monolithe `backend/main.py`.**
+  `PUT .../save|xlsx/save`, `DELETE/POST/PATCH /api/file`, `POST/PATCH/DELETE
+  /api/directory`, `POST /api/move`, `POST .../batch-upload` sont servis par
+  le nouveau `backend/routers/files_write.py` (effets de bord inchangés :
+  audit, index, SSE, webhooks, plugins, historique) ; 15 modèles dans
+  `schemas.py`.
+
+## [2.27.7] — 2026-09-26
+
+### Modifié
+
+- **#85 (T6a) — extraction lecture fichiers hors du monolithe `backend/main.py`.**
+  `/api/browse/{vault}`, `/api/file/{vault}/raw|download|backlinks` et
+  `GET /api/file/{vault}` (vue rendue tous formats) sont servis par le
+  nouveau `backend/routers/files_read.py` ; modèles dans `schemas.py`,
+  `_content_disposition`/`_media_max_inline_bytes` dans
+  `backend/routers/helpers.py` (partagés avec les tranches suivantes).
+  Correctif au passage : décorateur orphelin `/s/{token}` resté en T3 et
+  double-enregistrement de `/api/conflicts` supprimés.
+
+## [2.27.6] — 2026-09-26
+
+### Modifié
+
+- **#85 (T5) — extraction du domaine `search` hors du monolithe `backend/main.py`.**
+  Les 11 routes (`/api/search`, `/advanced`, `/replace`, `/tags`,
+  `/tree-search`, `/vault/{vault}/paths`, `/suggest`, `/tags/suggest`,
+  `/graph/{vault}`, `/index/reload`, `/index/reload/{vault}`) sont servies
+  par le nouveau `backend/routers/search.py` ; les modèles search dans
+  `schemas.py` et le pool de threads dans `backend/search_executor.py`
+  (même dimensionnement, même cycle de vie) — comportement inchangé.
+
+## [2.27.5] — 2026-09-26
+
+### Modifié
+
+- **#85 (T4) — extraction du domaine `backups` hors du monolithe `backend/main.py`.**
+  Les 9 routes (`/api/file/{vault}/backups|diff|restore`, `/api/backups`,
+  `/delete`, `/purge`, `/content`, `/compress`, `/auto`) sont servies par le
+  nouveau `backend/routers/backups.py` ; `Diff/Restore*` déménagent dans
+  `schemas.py` et le singleton SSE dans `backend/sse.py` (partagé avec
+  `main`) — comportement inchangé, aucun impact utilisateur.
+
+## [2.27.4] — 2026-09-26
+
+### Modifié
+
+- **#85 (T3) — extraction du domaine `sharing` hors du monolithe `backend/main.py`.**
+  `POST /api/share/{vault}`, `GET /api/shares`, `DELETE /api/share/{share_id}`
+  et les pages publiques `/s/{token}`, `/s/{token}/raw`, `/s/{token}/pdf`
+  sont servis par le nouveau `backend/routers/sharing.py` — chemins,
+  réponses, tags OpenAPI et authentification inchangés (aucun impact
+  utilisateur).
+
+## [2.27.3] — 2026-09-26
+
+### Modifié
+
+- **#85 (T2) — extraction du domaine `webhooks` hors du monolithe `backend/main.py`.**
+  Le CRUD `GET/POST/PATCH/DELETE /api/webhooks` (admin) est servi par le
+  nouveau `backend/routers/webhooks.py` — chemins, réponses, tags OpenAPI et
+  authentification inchangés (aucun impact utilisateur).
+
+## [2.27.2] — 2026-09-26
+
+### Modifié
+
+- **#85 (T1) — extraction du domaine `health` hors du monolithe `backend/main.py`.**
+  `GET /api/health` et `GET /api/health/detailed` (admin) sont servis par le
+  nouveau `backend/routers/health.py` (monté dans `main.py`) et le modèle
+  `HealthResponse` déménage dans `backend/schemas.py` — chemins, réponses,
+  tags OpenAPI et authentification inchangés (aucun impact utilisateur).
+
+---
+
+## [2.27.1] — 2026-09-26
+
+### Modifié
+
+- **Roadmap — priorisation dette & sécurité (décisions 2026-09-26).**
+  Items #85 (refonte architecturale) et #87 (CI/CD) détaillés et marqués
+  prioritaires : `backend/main.py` mesuré à ~4 827 lignes, `tools/registry.py`
+  à créer, persistance SQLite/Redis, verrous asyncio, audit des `except`
+  larges, CI sécurité bloquante (bandit/semgrep/trivy, audits pip/npm),
+  finition CSP nonce (BUG-034), cookies `Secure` par défaut, rotation clé
+  DeepSeek à confirmer (BUG-006). #73 Sync reporté (P4, hors chemin
+  critique) ; desktop #77 confirmé non signé + doc SmartScreen, reste les
+  6 tests E2E manuels. Corrections : sections livrées #83/#84 retirées du
+  backlog (détail dans l'archive, index inchangé), total restant recalculé
+  (~12-18 jours chemin critique : #77 fin + #85 + #87).
+
+---
+
+## [2.27.0] — 2026-09-25
+
+### Ajouté
+
+- **#152 — Viewer XLSX** : affichage des fichiers `.xlsx` en tableaux multi-feuilles (onglets,
+  en-têtes A1), édition inline des cellules avec `PUT /api/file/{vault}/xlsx/save` (backup avant
+  écriture, coercion numérique) et téléchargement du fichier d'origine.
+
+---
+
+## [2.25.1] — 2026-09-24
+
+### Corrigé
+
+- **`/api/diagnostics` — erreur 500 « dictionary changed size during iteration ».**
+  Le calcul des statistiques d'index itérait `inv.word_index` et `index` en
+  direct, pendant que l'indexeur les modifiait depuis un autre thread (build au
+  démarrage, hooks incrémentaux) : l'itérateur de dict levait `RuntimeError` et
+  l'endpoint renvoyait 500. Les deux dicts sont désormais **copiés avant
+  itération** (copie atomique sous le GIL), ce qui supprime la course.
+
+---
+
+## [2.25.0] — 2026-09-24
+
+### Ajouté
+
+- **#115 — barre d'outils de lecture toujours visible.** La barre d'actions d'un
+  document (pop-out, bookmark, Editer, Source, Copier, Partager…) est désormais
+  **épinglée en haut** de la zone de lecture : elle reste accessible en permanence,
+  même au bas d'un document long, au lieu de disparaître au défilement. Elle est
+  rendue comme un enfant direct du conteneur de défilement (`.file-toolbar`), masquée
+  en mode lecture, et alignée dans la fenêtre pop-out.
+- **#117 — avatars prédéfinis dans le profil.** La section **Profil** propose une
+  galerie de **12 avatars** fournis avec l'application (`frontend/icons/avatar/`) :
+  un clic charge l'image, la recadre au format carré 256 px (même pipeline que
+  l'import) et l'enregistre sur le compte. L'avatar actif est surligné ; l'import
+  d'une photo personnalisée et la suppression restent disponibles.
+
+### Corrigé
+
+- **BUG-078 — coloration syntaxique des fichiers de code perdue.** Les feuilles de
+  style highlight.js étaient basculées à partir de la **clé de thème**
+  (`defaut-obsigate`, …) au lieu du **mode** (`dark`/`light`) : les deux feuilles se
+  retrouvaient désactivées et les blocs de code (`.py`, `.sh`, `.ps1`, `.yml`, JSON,
+  blocs Markdown…) s'affichaient en texte brut. Le basculement est désormais piloté
+  par le mode, de façon déterministe, dans le moteur de thème (`themes.js`) comme au
+  premier rendu (`ui.js`) ; sépia et contraste élevé réutilisent la palette claire.
+
+---
+
+## [2.24.0] — 2026-09-24
+
+### Ajouté
+
+- **BUG-077 — assistant IA : bouton « Stop ».** Pendant qu'une réponse se diffuse, le
+  bouton d'envoi du composeur devient un bouton **Stop** (icône carrée, couleur
+  d'alerte) : un clic interrompt immédiatement l'exécution de l'agent (abort de la
+  requête SSE, annulation de la tâche côté serveur à la déconnexion) et la réponse
+  partielle est conservée avec un marqueur « ⏹ Exécution arrêtée. ». Le bouton reste
+  actif (il n'est plus désactivé) tant que l'agent travaille, y compris pendant la
+  reprise d'une confirmation.
+
+### Modifié
+
+- **BUG-075 — assistant IA : approbation globale des actions.** Une même réponse du
+  modèle peut demander **plusieurs** mutations (créer un dossier et les fichiers
+  qu'il contient…). Elles sont désormais **regroupées en une seule confirmation** (au
+  lieu d'une action après l'autre) : la carte liste chaque action avec son libellé et
+  son aperçu de diff, et un unique bouton « **Tout approuver (N)** » envoie
+  `confirm_all` — les actions en attente sont appliquées d'un bloc, puis la suite de
+  l'exécution est autorisée sans nouvelle carte. Les appels en lecture du même lot
+  s'exécutent immédiatement (conversation valide). Côté backend, `pending.actions`
+  remplace le report `deferred` des mutations d'un lot (`backend/agent/loop.py`) et
+  `confirm_all` arme `ToolContext.confirmed` pour le reste du run
+  (`backend/bookslm_routes.py`).
+
+### Corrigé
+
+- **BUG-074 — assistant IA : bloc d'étapes sans titre et compteur figé à 1.** Le
+  résumé replié affiche désormais un **titre** (le libellé de la première action,
+  « N étapes — Fichier créé : notes/a.md ▶ »), le compteur ne compte plus les
+  **réflexions** (seules les actions) et la reprise d'une confirmation continue de
+  diffuser dans **le même message** au lieu de créer un nouveau message « 1 étape »
+  par approbation : le bloc d'étapes s'accumule sur tout l'échange.
+- **BUG-076 — assistant IA : arborescence et document ouvert non rafraîchis.** Après
+  une action mutatrice de l'agent (création/suppression/renommage de fichier ou
+  dossier, écriture), l'arborescence est rafraîchie immédiatement (rafraîchissement
+  débouncé sur les événements `tool` mutateurs, sans attendre le watcher) et le
+  document affiché est rechargé depuis le disque ; les outils de création de documents
+  (xlsx/docx/csv/pdf) notifient désormais aussi la visionneuse.
+
+---
+
+## [2.23.0] — 2026-09-24
+
+### Ajouté
+
+- **#114 — Configuration, refonte mobile-responsive.** La page « Configurations »
+  est désormais pensée pour le tactile (≤ 768 px) : modale **plein écran**
+  (`100dvh`, safe-area), sommaire en **drawer coulissant** gauche
+  (`position: fixed`, largeur `min(320px, 88vw)`) avec fond assombri
+  (`#config-modal.config-toc-open::before`) et bouton de fermeture
+  `#config-toc-close` (tap sur le backdrop ou Échap ferme le drawer d'abord,
+  puis la modale) ; formulaires sur **une colonne** ; tous les boutons
+  (save/secondary/danger/add/sm/hamburger/fermer) et liens du sommaire en cibles
+  **≥ 44 px** ; champs et selects en **16 px + min-height 44 px** (anti-zoom
+  iOS) ; rangée « Sauvegarder / Réindexer / Réinitialiser » **sticky** en bas de
+  sa section avec safe-area ; MFA (codes de secours en 1 colonne, champ de code
+  full-width et wrap des rangées d'action) ; débordements webhook/token/share
+  corrigés. Dettes HTML/i18n de l'audit incluses : `.config-actions-row` replacée
+  dans `#cfg-backend-settings` (+ `</section>` orphelin supprimé), id dupliqué
+  `cfg-partages-publics` retiré du `<h2>`, conteneur mort
+  `#plugins-settings-container` supprimé, libellé `#mt-explorer` i18n
+  (`settings.explorer`), doublons `.config-btn-add` et règle morte
+  `.mfa-recovery-input` purgés ; i18n : clés mortes `settings.backend`,
+  `settings.backend_hint`, `settings.restart_badge`, `settings.save`,
+  `settings.plugins` supprimées, `config.toc_close` ajoutée, `settings.tabs` FR
+  corrigé (« Onglets »). Tests : `config-mobile.test.mjs` étendu (27, au CI) +
+  E2E `config-mobile.spec.js` (5, projet `chromium-mobile`). Fiche :
+  [docs/features/settings-mobile-114.md](./docs/features/settings-mobile-114.md).
+
+---
+
+## [2.22.1] — 2026-09-23
+
+### Corrigé
+
+- **BUG-073 — mobile, fin de contenu masquée.** La barre de navigation fixe du bas
+  (`#mobile-toolbar`, 64 px + safe-area) recouvrait le bas de tous les documents et
+  pages en vue mobile (≤ 768 px) : les dernières lignes restaient définitivement sous
+  la barre. Cause : la règle de clearance du bloc mobile ciblait `.main-layout`,
+  classe absente de `index.html` (le vrai conteneur est `.main-body`) — sélecteur
+  mort, aucun dégagement réservé. Correctif : cible `.main-body`
+  (`padding-bottom: calc(64px + env(safe-area-inset-bottom, 0))`), suppression de la
+  règle sœur morte `.editor-modal.active ~ .main-layout`, reset du dégagement en mode
+  lecture (barre masquée) et `body.np-active .content-area` ramené à `76 px` (le
+  dégagement dock, sans double comptage avec `.main-body`). Tests :
+  `tests/frontend/mobile-toolbar.test.mjs` (7, au CI) et E2E
+  `tests/e2e/mobile-toolbar.spec.js` (2, projet `chromium-mobile`).
+
+---
+
 ## [2.22.0] — 2026-09-23

 ### Ajouté
@@ -4,7 +4,7 @@

 **Porte d'entrée web ultra-léger pour vos vaults Obsidian** — Accédez, naviguez et recherchez dans toutes vos notes Obsidian depuis n'importe quel appareil via une interface web moderne et responsive.

-[![Version](https://img.shields.io/badge/Version-2.22.0-blue.svg)]()
+[![Version](https://img.shields.io/badge/Version-2.33.0-blue.svg)]()
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 [![Docker](https://img.shields.io/badge/Docker-Ready-blue.svg)](https://www.docker.com/)
 [![Python](https://img.shields.io/badge/Python-3.11+-green.svg)](https://www.python.org/)
@@ -23,7 +23,7 @@ Les **guides d'utilisation** pas à pas se trouvent dans [`docs/GUIDES/`](docs/G
 | Guide | Contenu |
 |---|---|
 | 🚀 [Prise en main](docs/GUIDES/PRISE_EN_MAIN.md) | Premier lancement, interface, navigation, vaults, raccourcis |
-| 🔍 [Recherche, PDF & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md) | Syntaxe de requête, recherche sémantique, lecteur PDF, diagrammes |
+| 🔍 [Recherche, PDF, Excel & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md) | Syntaxe de requête, recherche sémantique, lecteurs PDF/Excel, diagrammes |
 | 🤖 [Assistant IA & Forge](docs/GUIDES/ASSISTANT_IA_FORGE.md) | Fournisseurs, éditeur IA, BooksLM, Forge, commandes `@` / `/` |
 | 📝 [Édition & collaboration](docs/GUIDES/COLLABORATION.md) | Édition simultanée, curseurs distants, persistance |
 | 📱 [PWA & hors-ligne](docs/GUIDES/PWA_HORS_LIGNE.md) | Installation, cache hors-ligne, file de synchro, notifications |
@@ -85,6 +85,7 @@ Les **guides d'utilisation** pas à pas se trouvent dans [`docs/GUIDES/`](docs/G
 - **🖼️ Images Obsidian** : Support complet des syntaxes d'images Obsidian avec résolution intelligente
 - **🎬 Audio & vidéo** : Lecteurs HTML5 intégrés (`.mp3 .wav .flac .mp4 .webm`…) avec streaming HTTP Range (lecture, déplacement, plein écran) et **lecture persistante** (mini-lecteur flottant / mini-fenêtre vidéo, retour au média ou arrêt à tout moment, contrôles écran verrouillé via Media Session), repli téléchargement si le format n'est pas lisible par le navigateur
 - **🎨 Diagrammes Excalidraw** : Visualiseur/éditeur natif des fichiers `.excalidraw` et `.excalidraw.md` (iframe sandboxée, auto-save, thème clair/sombre, texte des diagrammes indexé pour la recherche)
+- **📊 Tableurs Excel** : les fichiers `.xlsx` s'ouvrent dans un visualiseur dédié — un tableau par feuille avec onglets, en-têtes A1 et édition directe des cellules (`PUT /api/file/{vault}/xlsx/save`, backup automatique, écriture atomique), plus le téléchargement du fichier d'origine. Les classeurs contenant des éléments qu'ObsiGate ne peut pas conserver (valeurs calculées, segments, contrôles de formulaire, signature…) affichent un **avertissement** et demandent confirmation avant l'enregistrement ; une saisie commençant par `=` ou `@` est stockée comme texte sauf activation du bouton `f(x)`
 - **🎨 Syntax highlight** : Coloration syntaxique des blocs de code
 - **🌓 Thème clair/sombre** : Toggle persisté en localStorage
 - **📡 Synchronisation temps réel** : Surveillance automatique des fichiers via watchdog avec mise à jour incrémentale de l'index
@@ -672,7 +673,7 @@ curl "http://localhost:2020/api/file/Recettes?path=pizza.md"

 ## 🔍 Recherche avancée

-> 📖 Guide complet : [Recherche, PDF & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md)
+> 📖 Guide complet : [Recherche, PDF, Excel & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md)

 ### Syntaxe de requête

@@ -975,8 +976,8 @@ Ce projet est sous licence **MIT** — voir le fichier [LICENSE](LICENSE) pour l

 ## 📝 Changelog

-Consultez le [CHANGELOG.md](./CHANGELOG.md) pour l'historique complet de toutes les versions (v1.0.0 → v2.22.0).
+Consultez le [CHANGELOG.md](./CHANGELOG.md) pour l'historique complet de toutes les versions (v1.0.0 → v2.33.0).

 ---

-*Projet : ObsiGate | Version : 2.22.0 | Dernière mise à jour : Septembre 2026*
+*Projet : ObsiGate | Version : 2.33.0 | Dernière mise à jour : Septembre 2026*
@@ -2,7 +2,7 @@

 **Ultra-light web gateway for your Obsidian vaults** — Access, browse, and search all your Obsidian notes from any device via a modern, responsive web interface.

-[![Version](https://img.shields.io/badge/Version-2.22.0-blue.svg)]()
+[![Version](https://img.shields.io/badge/Version-2.33.0-blue.svg)]()
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 [![Docker](https://img.shields.io/badge/Docker-Ready-blue.svg)](https://www.docker.com/)
 [![Python](https://img.shields.io/badge/Python-3.11+-green.svg)](https://www.python.org/)
@@ -21,7 +21,7 @@ Step-by-step **user guides** live in [`docs/GUIDES/`](docs/GUIDES/):
 | Guide | What it covers |
 |---|---|
 | 🚀 [Getting Started](docs/GUIDES/PRISE_EN_MAIN.md) | First run, interface, navigation, vaults, shortcuts |
-| 🔍 [Search, PDF & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md) | Query syntax, semantic search, PDF viewer, diagrams |
+| 🔍 [Search, PDF, Excel & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md) | Query syntax, semantic search, PDF/Excel viewers, diagrams |
 | 🤖 [AI Assistant & Forge](docs/GUIDES/ASSISTANT_IA_FORGE.md) | Providers, AI editor, BooksLM, Forge, `@` / `/` commands |
 | 📝 [Editing & Collaboration](docs/GUIDES/COLLABORATION.md) | Simultaneous editing, remote cursors, persistence |
 | 📱 [PWA & Offline](docs/GUIDES/PWA_HORS_LIGNE.md) | Install as an app, offline cache, sync queue, push |
@@ -84,6 +84,7 @@ Step-by-step **user guides** live in [`docs/GUIDES/`](docs/GUIDES/):
 - **🖼️ Obsidian Images** : Full support for all Obsidian image syntaxes with intelligent resolution
 - **🎬 Audio & video** : Built-in HTML5 players (`.mp3 .wav .flac .mp4 .webm`…) with HTTP Range streaming (play, seek, fullscreen) and **persistent playback** (floating mini-player / mini video window, return to media or stop anytime, lock-screen controls via Media Session), falling back to download when the format is not playable in the browser
 - **🎨 Excalidraw Diagrams** : Native viewer/editor for `.excalidraw` and `.excalidraw.md` files (sandboxed iframe, autosave, dark/light theme, diagram text indexed for search)
+- **📊 Excel Spreadsheets** : `.xlsx` files open in a dedicated viewer — one table per sheet with tabs, A1 headers and inline cell editing (`PUT /api/file/{vault}/xlsx/save`, automatic backup, atomic write), plus download of the original file. Workbooks holding elements ObsiGate cannot preserve (cached values, slicers, form controls, signature…) show a **warning** and ask for confirmation before saving; a value starting with `=` or `@` is stored as text unless the `f(x)` toggle is enabled
 - **🎨 Syntax Highlight** : Syntax highlighting for code blocks
 - **🌓 Light/Dark Theme** : Toggle persisted in localStorage
 - **📡 Real-time Sync** : Automatic file monitoring via watchdog with incremental index updates
@@ -803,7 +804,7 @@ curl "http://localhost:2020/api/file/Recipes?path=pizza.md"

 ## 🔍 Advanced Search

-> 📖 Full guide: [Search, PDF & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md)
+> 📖 Full guide: [Search, PDF, Excel & Excalidraw](docs/GUIDES/RECHERCHE_PDF_EXCALIDRAW.md)

 ### Query Syntax

@@ -1150,8 +1151,8 @@ This project is licensed under the **MIT License** - see the [LICENSE](LICENSE)

 ## 📝 Changelog

-See [CHANGELOG.md](./CHANGELOG.md) for the complete version history (v1.0.0 → v2.22.0).
+See [CHANGELOG.md](./CHANGELOG.md) for the complete version history (v1.0.0 → v2.33.0).

 ---

-*Project: ObsiGate | Version: 2.22.0 | Last updated: September 2026*
+*Project: ObsiGate | Version: 2.33.0 | Last updated: September 2026*
@@ -1 +1 @@
-2.22.0
+2.33.0
@@ -28,6 +28,7 @@ from backend.tools.api import (
    ToolError,
    ToolScope,
    call_tool,
+    get_tool,
    get_tool_schemas,
 )
 from backend.tools.labels import thought_step_label, tool_step_label
@@ -121,12 +122,15 @@ def _assistant_tool_message(content: str | None, tool_calls: list[Any]) -> dict[
 def _deferred_tool_message(call: Any, reason: str | None = None) -> dict[str, Any]:
    """Answer a tool call that was not reached because the run stopped early.

-    A single LLM response may carry several tool calls. When one of them is
-    mutating and pauses the run for confirmation, the assistant message already
-    lists *all* of them, so every ``tool_call_id`` must get a tool result before
-    the next LLM call (the OpenAI tool protocol rejects dangling ids). The calls
-    that were not reached get a synthetic ``deferred`` result; the model
-    re-issues them once the confirmed call has been applied (BUG-050).
+    A single LLM response may carry several tool calls; when the run stops
+    before reaching some of them (tool-call quota), the assistant message still
+    lists *all* of them, so every ``tool_call_id`` must get a tool result
+    before the next LLM call (the OpenAI tool protocol rejects dangling ids).
+    The calls that were not reached get a synthetic ``deferred`` result.
+
+    Note: mutating calls that pause the run for confirmation are no longer
+    deferred — they are batched and applied together on resume (BUG-075); this
+    helper remains for budget stops (BUG-050/BUG-052).
    """
    return {
        "role": "tool",
@@ -135,13 +139,29 @@ def _deferred_tool_message(call: Any, reason: str | None = None) -> dict[str, An
        "content": json.dumps({
            "status": "deferred",
            "reason": reason or (
-                "Not executed: the run paused to confirm an earlier tool call. "
+                "Not executed: the run stopped before reaching this tool call. "
                "Re-issue this call if it is still needed."
            ),
        }, ensure_ascii=False),
    }


+def _action_descriptor(call: Any) -> dict[str, Any]:
+    """Describe one paused mutating tool call for the confirmation payload.
+
+    A single LLM response may request several mutations (create a folder and
+    the files inside it…). They are batched into one confirmation so the user
+    approves the whole plan in one click (BUG-075). ``step`` reuses the
+    Notion-style label, so the confirmation card reads like the steps block.
+    """
+    return {
+        "id": call.id,
+        "tool": call.name,
+        "arguments": call.arguments,
+        "step": tool_step_label(call.name, call.arguments),
+    }
+
+
 def _fallback_summary(executed: list[ToolCallRecord]) -> str:
    """Deterministic non-empty answer built from the gathered tool results.

@@ -212,53 +232,66 @@ def _execute_confirmed(
    executed: list[ToolCallRecord],
    on_tool_call: Callable[[ToolCallRecord], None] | None,
 ) -> None:
-    """Apply a previously-paused mutating tool call and feed its result back.
+    """Apply previously-paused mutating tool calls and feed their results back.

    The pending payload is the ``error`` object emitted by a ``confirmation``
-    event. The assistant tool-call message is expected to already be in
+    event, optionally carrying an ``actions`` list with every mutating call of
+    the LLM turn (BUG-075). Each action is applied with a one-shot confirmation
+    and its ``tool_call_id`` answered, keeping the conversation valid for the
+    resumed turn. The assistant tool-call message is expected to already be in
    ``convo`` (it is part of the snapshot returned with the confirmation).
    """
    from backend.ai_chat import ToolCall

-    error = confirm_pending.get("error", confirm_pending)
-    name = error.get("tool")
-    arguments = error.get("arguments") or {}
-    call_id = error.get("id") or "call_pending"
+    error = confirm_pending.get("error", confirm_pending) or {}
+    actions = confirm_pending.get("actions")
+    if not isinstance(actions, list) or not actions:
+        # Legacy single-action payload (no ``actions`` list).
+        actions = [{
+            "id": error.get("id") or "call_pending",
+            "tool": error.get("tool"),
+            "arguments": error.get("arguments") or {},
+        }]

-    if not name:
-        raise ToolError("Malformed confirmation payload", code="invalid_confirmation")
+    for action in actions:
+        name = action.get("tool")
+        arguments = action.get("arguments") or {}
+        call_id = action.get("id") or "call_pending"

-    # Make sure the assistant tool-call message is present in the snapshot.
-    if not any(
-        m.get("role") == "assistant" and any(
-            tc.get("id") == call_id for tc in (m.get("tool_calls") or [])
+        if not name:
+            raise ToolError("Malformed confirmation payload", code="invalid_confirmation")
+
+        # Make sure the assistant tool-call message is present in the snapshot.
+        if not any(
+            m.get("role") == "assistant" and any(
+                tc.get("id") == call_id for tc in (m.get("tool_calls") or [])
+            )
+            for m in convo
+        ):
+            convo.append(_assistant_tool_message(None, [ToolCall(id=call_id, name=name, arguments=arguments)]))
+
+        try:
+            result = call_tool(name, ctx, arguments, confirm=True)
+            payload = result.data
+            ok = True
+        except ToolError as e:
+            payload = e.to_dict()
+            ok = False
+
+        record = ToolCallRecord(
+            name=name, arguments=arguments, ok=ok, result=payload,
+            step=tool_step_label(name, arguments),
        )
-        for m in convo
-    ):
-        convo.append(_assistant_tool_message(None, [ToolCall(id=call_id, name=name, arguments=arguments)]))
+        executed.append(record)
+        if on_tool_call is not None:
+            on_tool_call(record)

-    try:
-        result = call_tool(name, ctx, arguments, confirm=True)
-        payload = result.data
-        ok = True
-    except ToolError as e:
-        payload = e.to_dict()
-        ok = False
-
-    record = ToolCallRecord(
-        name=name, arguments=arguments, ok=ok, result=payload,
-        step=tool_step_label(name, arguments),
-    )
-    executed.append(record)
-    if on_tool_call is not None:
-        on_tool_call(record)
-
-    convo.append({
-        "role": "tool",
-        "tool_call_id": call_id,
-        "name": name,
-        "content": json.dumps(_truncate(payload), ensure_ascii=False, default=str),
-    })
+        convo.append({
+            "role": "tool",
+            "tool_call_id": call_id,
+            "name": name,
+            "content": json.dumps(_truncate(payload), ensure_ascii=False, default=str),
+        })


 async def run_agent(
@@ -315,6 +348,36 @@ async def run_agent(
    convo = [dict(m) for m in (resume_messages if resume_messages is not None else messages)]
    executed: list[ToolCallRecord] = []

+    def _run_call(call: Any) -> None:
+        """Execute one tool call, record it and answer its ``tool_call_id``.
+
+        ``ToolConfirmationRequired`` propagates to the caller so the loop can
+        pause and batch the mutating calls of the turn (BUG-075).
+        """
+        try:
+            result = call_tool(call.name, ctx, call.arguments)
+            payload: Any = result.data
+            ok = True
+        except ToolConfirmationRequired:
+            raise
+        except ToolError as e:
+            payload = e.to_dict()
+            ok = False
+        record = ToolCallRecord(
+            name=call.name, arguments=call.arguments, ok=ok, result=payload,
+            step=tool_step_label(call.name, call.arguments),
+        )
+        executed.append(record)
+        steps.append(record.step)
+        if on_tool_call is not None:
+            on_tool_call(record)
+        convo.append({
+            "role": "tool",
+            "tool_call_id": call.id,
+            "name": call.name,
+            "content": json.dumps(_truncate(payload), ensure_ascii=False, default=str),
+        })
+
    if confirm_pending:
        if quota is not None and len(executed) >= quota:
            return AgentResult(
@@ -357,19 +420,25 @@ async def run_agent(
                    llm, convo, executed, steps, iteration, STOP_QUOTA_EXCEEDED
                )
            try:
-                result = call_tool(call.name, ctx, call.arguments)
-                payload = result.data
-                ok = True
+                _run_call(call)
            except ToolConfirmationRequired as e:
                logger.info(f"Agent paused: confirmation required for '{call.name}'")
                pending = e.to_dict()
                # Include the tool-call id so the client can echo it back.
                pending["error"]["id"] = call.id
-                # BUG-050: the assistant message lists every tool call of this
-                # batch, so answer the ones we did not reach to keep the
-                # conversation valid for the resumed turn.
-                for skipped in response.tool_calls[index + 1:]:
-                    convo.append(_deferred_tool_message(skipped))
+                # BUG-075: batch every mutating call of this LLM turn so the
+                # user approves the whole plan at once (one resume applies them
+                # all) instead of approving one action after another. Read-only
+                # calls of the batch run immediately and answer their
+                # ``tool_call_id`` so the resumed turn stays valid.
+                actions = [_action_descriptor(call)]
+                for after in response.tool_calls[index + 1:]:
+                    spec = get_tool(after.name)
+                    if spec is not None and spec.requires_confirmation:
+                        actions.append(_action_descriptor(after))
+                    else:
+                        _run_call(after)
+                pending["actions"] = actions
                return AgentResult(
                    content=response.content or "",
                    messages=convo,
@@ -379,25 +448,6 @@ async def run_agent(
                    stopped=STOP_CONFIRMATION_REQUIRED,
                    pending=pending,
                )
-            except ToolError as e:
-                payload = e.to_dict()
-                ok = False
-
-            record = ToolCallRecord(
-                name=call.name, arguments=call.arguments, ok=ok, result=payload,
-                step=tool_step_label(call.name, call.arguments),
-            )
-            executed.append(record)
-            steps.append(record.step)
-            if on_tool_call is not None:
-                on_tool_call(record)
-
-            convo.append({
-                "role": "tool",
-                "tool_call_id": call.id,
-                "name": call.name,
-                "content": json.dumps(_truncate(payload), ensure_ascii=False, default=str),
-            })

    logger.warning(f"Agent reached max iterations ({max_iterations})")
    return await _finalize_answer(
@@ -119,25 +119,30 @@ def decode_token(token: str) -> dict | None:
 _revoked_map: dict[str, int] = {}
 _revoked_loaded = False

+# ROADMAP #85 T10a — verrou autour du read-modify-write du store de
+# révocation (perte de révocations en cas de logouts concurrents).
+_revoked_lock = threading.RLock()
+

 def _load_revoked():
    """Load revoked token JTIs from disk into memory (once)."""
    global _revoked_loaded, _revoked_map
-    if _revoked_loaded:
-        return
-    if REVOKED_TOKENS_FILE.exists():
-        try:
-            data = json.loads(REVOKED_TOKENS_FILE.read_text())
-            # Drop entries whose underlying token has itself expired.
-            now = int(time.time())
-            _revoked_map = {
-                jti: int(exp) for jti, exp in data.items()
-                if int(exp) > now
-            }
-        except Exception as e:
-            logger.warning(f"Failed to load revoked tokens: {e}")
-            _revoked_map = {}
-    _revoked_loaded = True
+    with _revoked_lock:
+        if _revoked_loaded:
+            return
+        if REVOKED_TOKENS_FILE.exists():
+            try:
+                data = json.loads(REVOKED_TOKENS_FILE.read_text())
+                # Drop entries whose underlying token has itself expired.
+                now = int(time.time())
+                _revoked_map = {
+                    jti: int(exp) for jti, exp in data.items()
+                    if int(exp) > now
+                }
+            except Exception as e:
+                logger.warning(f"Failed to load revoked tokens: {e}")
+                _revoked_map = {}
+        _revoked_loaded = True


 def _save_revoked():
@@ -154,24 +159,26 @@ def revoke_token(jti: str, expires_at: int | None = None):
    ``expires_at`` is the revoked token's own ``exp`` (unix seconds) — the
    record is kept at least that long so a long-lived API token cannot
    outlive its revocation. ``None`` means the token never expires (API/MCP
-    "sans fin") → the record is kept forever (capped at ~100 years, the JWT
+    "sans fin") → the     record is kept forever (capped at ~100 years, the JWT
    store's practical infinity). Default keeps 7 days (session tokens).
    """
-    _load_revoked()
-    now = int(time.time())
-    if expires_at is None:
-        until = now + 100 * 365 * 24 * 3600
-    else:
-        until = max(int(expires_at), now + REFRESH_TOKEN_EXPIRE_SECONDS)
-    _revoked_map[jti] = until
-    _save_revoked()
+    with _revoked_lock:
+        _load_revoked()
+        now = int(time.time())
+        if expires_at is None:
+            until = now + 100 * 365 * 24 * 3600
+        else:
+            until = max(int(expires_at), now + REFRESH_TOKEN_EXPIRE_SECONDS)
+        _revoked_map[jti] = until
+        _save_revoked()
    logger.debug(f"Revoked token JTI: {jti[:8]}...")


 def is_token_revoked(jti: str) -> bool:
    """Check if a token JTI has been revoked."""
-    _load_revoked()
-    return jti in _revoked_map
+    with _revoked_lock:
+        _load_revoked()
+        return jti in _revoked_map


 # ---------------------------------------------------------------------------
@@ -5,6 +5,7 @@
 import base64
 import binascii
 import logging
+import os
 import re

 from fastapi import APIRouter, Body, Depends, HTTPException, Request, Response
@@ -15,7 +16,7 @@ from backend.ratelimit import record_account_failure as rl_record_account_failur
 from backend.ratelimit import record_account_success as rl_record_account_success
 from backend.ratelimit import record_failure as rl_record_failure
 from backend.ratelimit import record_success as rl_record_success
-from backend.services.net import get_client_ip
+from backend.services.net import get_client_ip, is_trusted_proxy

 from .jwt_handler import (
    ACCESS_TOKEN_EXPIRE_SECONDS,
@@ -56,6 +57,34 @@ logger = logging.getLogger("obsigate.auth.router")
 router = APIRouter(prefix="/api/auth", tags=["auth"])


+def is_secure_cookies(request: Request | None = None) -> bool:
+    """True when auth cookies must carry the ``Secure`` flag (#87 T3/T8).
+
+    ``OBSIGATE_SECURE_COOKIES=true|false|auto`` (défaut : ``auto``) :
+    ``true``/``false`` forcent le comportement ; ``auto`` met ``Secure``
+    si la requête arrive en https (production derrière TLS) et l'omet
+    sinon (dev local en http — les navigateurs jettent les cookies
+    ``Secure`` sur http, ce qui casserait silencieusement les logins
+    localhost). Derrière un reverse proxy qui termine TLS, le schéma perçu
+    est http : avec ``OBSIGATE_TRUST_PROXY=true``, ``X-Forwarded-Proto``
+    est honoré (même garde que ``get_client_ip``, BUG-030).
+    """
+    forced = os.environ.get("OBSIGATE_SECURE_COOKIES", "auto").lower()
+    if forced in ("1", "true", "yes", "on"):
+        return True
+    if forced in ("0", "false", "no", "off"):
+        return False
+    if request is None:
+        return False
+    if request.url.scheme == "https":
+        return True
+    if is_trusted_proxy():
+        proto = request.headers.get("x-forwarded-proto", "").split(",")[0].strip().lower()
+        if proto == "https":
+            return True
+    return False
+
+
 # ── Pydantic request models ──────────────────────────────────────────

 class LoginRequest(BaseModel):
@@ -218,10 +247,11 @@ async def login(body: LoginRequest, response: Response, request: Request):
            "remember_me": body.remember_me,
        }

-    return _issue_tokens(user, body.username, body.remember_me, response)
+    return _issue_tokens(user, body.username, body.remember_me, response, request)


-def _issue_tokens(user: dict, username: str, remember_me: bool, response: Response) -> dict:
+def _issue_tokens(user: dict, username: str, remember_me: bool, response: Response,
+                  request: Request | None = None) -> dict:
    """Issue JWT tokens after successful authentication (password or MFA verified)."""
    record_login_success(username)
    rl_record_account_success(username)
@@ -229,9 +259,8 @@ def _issue_tokens(user: dict, username: str, remember_me: bool, response: Respon
    access_token = create_access_token(user)
    refresh_token, refresh_jti = create_refresh_token(username, remember=remember_me)

-    import os
    max_age = 2592000 if remember_me else 604800  # 30d or 7d
-    secure = os.environ.get("OBSIGATE_SECURE_COOKIES", "false").lower() == "true"
+    secure = is_secure_cookies(request)
    response.set_cookie(
        key="refresh_token",
        value=refresh_token,
@@ -253,7 +282,8 @@ def _issue_tokens(user: dict, username: str, remember_me: bool, response: Respon
    )
    return {
        "access_token": access_token,
-        "token_type": "bearer",  # nosec B105 — OAuth2 token_type, pas un mot de passe
+        # OAuth2 token_type, pas un mot de passe (B105) :
+        "token_type": "bearer",  # nosec B105
        "expires_in": ACCESS_TOKEN_EXPIRE_SECONDS,
        "user": {
            "username": user["username"],
@@ -299,9 +329,7 @@ async def refresh_token_endpoint(request: Request, response: Response):
        if stale:
            raise HTTPException(401, "Session expirée, veuillez vous reconnecter")

-    import os
-
-    secure = os.environ.get("OBSIGATE_SECURE_COOKIES", "false").lower() == "true"
+    secure = is_secure_cookies(request)
    remember_me = bool(payload.get("remember", False))

    # BUG-027: rotate the refresh token — the old one is now single-use.
@@ -332,7 +360,8 @@ async def refresh_token_endpoint(request: Request, response: Response):

    return {
        "access_token": new_access_token,
-        "token_type": "bearer",  # nosec B105 — OAuth2 token_type, pas un mot de passe
+        # OAuth2 token_type, pas un mot de passe (B105) :
+        "token_type": "bearer",  # nosec B105
        "expires_in": ACCESS_TOKEN_EXPIRE_SECONDS,
    }

@@ -426,6 +455,7 @@ async def patch_me(req: UpdateMeRequest, current_user=Depends(require_auth)):
 async def change_password(
    req: ChangePasswordRequest,
    response: Response,
+    request: Request,
    current_user=Depends(require_auth),
 ):
    """Change own password.
@@ -441,7 +471,7 @@ async def change_password(
    updated = get_user(current_user["username"])
    result: dict = {"message": "Mot de passe mis à jour"}
    if updated is not None:
-        result.update(_issue_tokens(updated, updated["username"], False, response))
+        result.update(_issue_tokens(updated, updated["username"], False, response, request))
    return result


@@ -804,7 +834,7 @@ async def mfa_webauthn_verify(

    rl_record_success(client_ip)
    logger.info(f"User '{body.username}' logged in via WebAuthn")
-    return _issue_tokens(user, body.username, body.remember_me, response)
+    return _issue_tokens(user, body.username, body.remember_me, response, request)


@router.get("/mfa/status")
@@ -812,6 +842,16 @@ async def mfa_status(current_user=Depends(require_auth)):
    """Return current user's MFA status."""
    from .user_store import get_user
    user = get_user(current_user["username"])
+    if user is None:
+        # BUG-081 : auth désactivée (OBSIGATE_AUTH_ENABLED=false) → le
+        # pseudo-user "anonymous" n'a aucune entrée en store : pas de MFA,
+        # et surtout pas de 500 (`AttributeError` sur `user.get`).
+        return {
+            "mfa_enabled": False,
+            "mfa_method": None,
+            "totp_enabled": False,
+            "webauthn_credentials": 0,
+        }
    return {
        "mfa_enabled": user.get("mfa_enabled", False),
        "mfa_method": user.get("mfa_method"),
@@ -847,7 +887,7 @@ async def mfa_totp_verify(body: MfaVerifyRequest, response: Response, request: R
    # Clear IP rate limit on success
    rl_record_success(client_ip)

-    return _issue_tokens(user, body.username, body.remember_me, response)
+    return _issue_tokens(user, body.username, body.remember_me, response, request)


@router.post("/mfa/recovery")
@@ -885,7 +925,7 @@ async def mfa_recovery_login(body: MfaRecoveryRequest, response: Response, reque
    rl_record_success(client_ip)

    logger.info(f"User '{body.username}' logged in via recovery code")
-    return _issue_tokens(user, body.username, False, response)
+    return _issue_tokens(user, body.username, False, response, request)


 # ── Admin endpoints ───────────────────────────────────────────────────
@@ -110,6 +110,12 @@ class BooksLMChatRequest(BaseModel):
        description="Conversation snapshot returned alongside a ``confirmation`` event, "
                    "echoed back to resume the agent run.",
    )
+    confirm_all: bool = Field(
+        default=False,
+        description="Global approval (BUG-075): apply every pending action of the batch "
+                    "and auto-approve the remaining mutating calls of the same run, "
+                    "so the run does not pause on each action.",
+    )
    app_context: dict[str, Any] | None = Field(
        default=None,
        description="Live client UI state for the General assistant: open_documents, "
@@ -506,9 +512,11 @@ async def api_bookslm_agent(
    Same context as ``/chat`` but the model may call tools (read/search the
    vault) through the shared tool layer. Emits one ``tool`` event per executed
    tool call, then a final ``message`` event. Mutating tools pause the run with
-    a ``confirmation`` event (two-step propose/apply) carrying the pending call
-    and the conversation snapshot; the client resumes by echoing them back in
-    ``confirm`` / ``confirm_messages``.
+    a ``confirmation`` event (two-step propose/apply) carrying the pending
+    ``actions`` (every mutating call of the turn) and the conversation snapshot;
+    the client resumes by echoing them back in ``confirm`` / ``confirm_messages``,
+    optionally with ``confirm_all`` to apply the whole batch and auto-approve the
+    rest of the run (BUG-075).
    """
    _validate_vision_support(req)
    system_prompt = _resolve_system_prompt(req, current_user, agent=True)
@@ -523,6 +531,10 @@ async def api_bookslm_agent(
    messages.append({"role": "user", "content": _build_user_content(req, vault_path)})

    ctx = ToolContext(user=current_user, mode=ToolMode.IN_APP)
+    if req.confirm_all:
+        # BUG-075: a single global approval authorizes the whole plan, so the
+        # run no longer pauses on every subsequent mutating call.
+        ctx.confirmed = True

    async def _llm(msgs, tool_schemas):
        return await chat_completion(
@@ -0,0 +1,35 @@
+"""Content-Security-Policy nonces (ROADMAP #87, tranche 5b).
+
+Chaque réponse HTTP reçoit un nonce frais (``request.state.csp_nonce``)
+injecté dans ``script-src``. Les routes servant du HTML avec des scripts
+inline (index, popout, admin, editor-poc, excalidraw, page de partage)
+l'injectent dans le balisage via :func:`inject_csp_nonce` — mêmes
+emplacements, aucun script déplacé.
+
+Tant que ``'unsafe-inline'`` reste dans la politique (retrait en T5c),
+l'injection est inerte : elle prépare la bascule sans changer le
+comportement.
+"""
+
+from __future__ import annotations
+
+import re
+import secrets
+
+# Balises <script> exécutables sans `src` et sans nonce existant :
+# `<script>`, `<script type="module">`, `<script type="importmap">`.
+# Les blocs non-JS (ex. `type="text/plain"`) et les scripts externes
+# (`src=…`, couverts par 'self'/hôtes CDN) sont laissés intacts.
+_SCRIPT_TAG_RE = re.compile(
+    r"<script(?=>|\s+type=\"(?:module|importmap)\"\s*>)",
+)
+
+
+def new_nonce() -> str:
+    """Generate a fresh per-response CSP nonce."""
+    return secrets.token_urlsafe(16)
+
+
+def inject_csp_nonce(html: str, nonce: str) -> str:
+    """Add ``nonce="…"`` to bare executable inline ``<script>`` tags."""
+    return _SCRIPT_TAG_RE.sub(f'<script nonce="{nonce}"', html)
@@ -23,6 +23,7 @@ import re
 import unicodedata
 import zipfile
 from pathlib import Path
+from typing import cast

 import frontmatter
 import mistune
@@ -246,7 +247,9 @@ def _render_body(md: str, file_dir: Path, vault_path: Path, current: Path) -> st
    """Render raw markdown to an HTML fragment (images inlined, wikilinks resolved)."""
    md = _inline_images(md, file_dir, vault_path)
    md = _convert_wikilinks(md, vault_path, current)
-    return _markdown(md)
+    # mistune 3.3 types `Markdown.__call__` as `str | list[...]` (le renderer
+    # HTML renvoie toujours `str` à l'exécution).
+    return cast(str, _markdown(md))


 def _build_nav(vault_path: Path, current: Path) -> str:
@@ -35,7 +35,8 @@ def diagram_png_for(code: str) -> Path | None:
    Mermaid, ou None. Le hash doit rester synchrone avec le script de build :
    sha1(unescape(code).strip())[:16]."""
    normalized = html.unescape(code).strip()
-    sha = hashlib.sha1(normalized.encode("utf-8")).hexdigest()[:16]
+    # Identifiant de cache déterministe (pas un usage sécurité).
+    sha = hashlib.sha1(normalized.encode("utf-8")).hexdigest()[:16]  # nosec B324
    png = DIAGRAMS_DIR / (sha + ".png")
    return png if png.exists() else None

@@ -65,6 +65,7 @@ SUPPORTED_EXTENSIONS = {
    ".sh", ".bash", ".zsh", ".fish", ".bat", ".cmd", ".ps1",
    ".json", ".yaml", ".yml", ".toml", ".xml", ".csv",
    ".cfg", ".ini", ".conf", ".env", ".pdf",
+    ".xlsx",
    ".html", ".css", ".scss", ".less",
    ".java", ".c", ".cpp", ".h", ".hpp", ".cs", ".go", ".rs", ".rb",
    ".php", ".sql", ".r", ".m", ".swift", ".kt",
@@ -350,6 +351,23 @@ def _decompress_excalidraw(compressed: str) -> dict[str, Any] | None:
    return data


+def extract_xlsx_indexable(file_path: Path) -> str:
+    """Return searchable text for a workbook (#153 A5).
+
+    Lazy wrapper: ``openpyxl`` is only imported when a spreadsheet is actually
+    indexed, so a vault without workbooks never pays the import. Errors are
+    swallowed — a corrupt or encrypted file still gets indexed by name.
+    """
+    try:
+        from backend.xlsx_reader import extract_indexable_text
+    except Exception:  # pragma: no cover - openpyxl missing
+        return ""
+    try:
+        return extract_indexable_text(file_path)
+    except Exception:  # pragma: no cover - defensive
+        return ""
+
+
 def extract_excalidraw_indexable(raw: str) -> str:
    """Return indexable text content for a raw .excalidraw / .excalidraw.md file.

@@ -559,6 +577,13 @@ def _scan_vault(
                    raw = ""
                    title = fpath.stem.replace("-", " ").replace("_", " ")
                    content_preview = ""
+                elif ext == ".xlsx":
+                    # #153 A5 — a workbook stays rendered by the viewer, but its
+                    # cell values are now indexed as text so a spreadsheet is
+                    # findable by its content (parity with _index_single_file_sync).
+                    raw = extract_xlsx_indexable(fpath)
+                    title = fpath.stem.replace("-", " ").replace("_", " ")
+                    content_preview = raw[:200].strip()
                else:
                    raw = fpath.read_text(encoding="utf-8", errors="replace")
                    title = fpath.stem.replace("-", " ").replace("_", " ")
@@ -800,6 +825,13 @@ async def reload_index() -> dict[str, Any]:
    await build_index()
    # BUG-040/#86: complete the deferred PDF + excalidraw extraction.
    await enrich_pdf_texts()
+    # The inverted index is NOT updated by the hooks here: the rebuild above
+    # replaces whole vault entries, so the incremental notifications are not
+    # emitted for the files that only changed content. Without this, a manual
+    # reindex left TF-IDF search serving a stale index (BUG-089).
+    from backend.search import init_inverted_index
+
+    init_inverted_index()
    stats = {}
    for name, data in index.items():
        stats[name] = {"file_count": len(data["files"]), "tag_count": len(data["tags"])}
@@ -875,6 +907,13 @@ async def reload_single_vault(vault_name: str) -> dict[str, Any]:
    # BUG-040/#86: complete the deferred PDF + excalidraw extraction.
    await enrich_pdf_texts(vault_name)

+    # Same as reload_index: the vault entry was replaced wholesale, so rebuild
+    # the inverted index or TF-IDF search keeps serving stale postings
+    # (BUG-089).
+    from backend.search import init_inverted_index
+
+    init_inverted_index()
+
    stats = {"file_count": len(vault_data["files"]), "tag_count": len(vault_data["tags"])}
    logger.info(f"Vault '{vault_name}' reindexed: {stats['file_count']} files, {stats['tag_count']} tags")
    return stats
@@ -947,6 +986,10 @@ def _index_single_file_sync(vault_name: str, vault_path: str, file_path: str, va
            # #108 — binary media: metadata only, never read the bytes.
            raw = ""
            content_preview = ""
+        elif ext == ".xlsx":
+            # #153 A5 — index sheet names + header rows as text (see _scan_vault).
+            raw = extract_xlsx_indexable(fpath)
+            content_preview = raw[:200].strip()
        else:
            raw = fpath.read_text(encoding="utf-8", errors="replace")
            content_preview = raw[:200].strip()
@@ -1215,6 +1258,12 @@ async def remove_vault_from_index(vault_name: str):
                if not _file_lookup[key]:
                    _file_lookup.pop(key, None)

+            # Notify the inverted index, otherwise every document of the vault
+            # stays in it as a ghost (postings, doc_info, doc_vault, vault_docs)
+            # and keeps matching searches for a vault that no longer exists.
+            if _on_index_change:
+                _on_index_change('remove', vault_name, rel_path, f)  # type: ignore[misc]
+
        # Clean path_index
        path_index.pop(vault_name, None)

@@ -32,7 +32,8 @@ def thumb_cache_path(file_path: Path, size: int) -> Path:
        stamp = f"{st.st_mtime_ns}:{st.st_size}"
    except OSError:
        stamp = "0:0"
-    key = hashlib.sha1(f"{file_path}:{stamp}:{size}".encode()).hexdigest()
+    # Clé de cache miniature (pas un usage sécurité).
+    key = hashlib.sha1(f"{file_path}:{stamp}:{size}".encode()).hexdigest()  # nosec B324
    return thumbs_cache_dir() / f"{key}.webp"


@@ -181,6 +181,30 @@ _ENDPOINT_EXAMPLES: dict[tuple[str, str], dict[str, Any]] = {
        "request": {"path": "notes/Accueil.md", "content": "# Accueil\n\nMis à jour."},
        "response": {"status": "ok", "vault": "TestVault", "path": "notes/Accueil.md", "size": 26},
    },
+    ("put", "/api/file/{vault_name}/xlsx/save"): {
+        "request": {"sheet": "Budget", "cells": {"B1": "250"}, "allow_formula": False, "force": False},
+        "response": {"status": "ok", "vault": "TestVault", "path": "data/budget.xlsx", "size": 1},
+    },
+    # GET : pas d'exemple de requête (un requestBody sur un GET serait un OpenAPI
+    # invalide) — les paramètres sont documentés par leurs Query().
+    ("get", "/api/file/{vault_name}/xlsx/sheet"): {
+        "response": {
+            "vault": "TestVault",
+            "path": "data/budget.xlsx",
+            "sheet": "Budget",
+            "offset": 0,
+            "limit": 200,
+            "rows": 2,
+            "cols": 2,
+            "total_rows": 640,
+            "total_cols": 12,
+            "max_rows": 500,
+            "max_cols": 40,
+            "truncated": True,
+            "has_more": True,
+            "html": "<table>…</table>",
+        },
+    },
    ("post", "/api/search/replace"): {
        "request": {"query": "Python", "replacement": "Python 3", "vault": "all", "dry_run": True},
        "response": {"matches": [{"vault": "TestVault", "path": "note1.md", "title": "Python", "match_count": 3}], "total_matches": 3, "dry_run": True},
@@ -12,14 +12,24 @@ the per-account lockout in ``user_store.py``.
   deployment, front this service with a shared store (Redis) or a single
   worker.  This limitation is intentional and documented (BUG-031).

+   Opt-in persistence (ROADMAP #85 T10b) : if ``OBSIGATE_RATELIMIT_DB`` points
+   to a SQLite file, counters are stored there instead (WAL mode, one short
+   connection per call — safe across threads, processes and restarts sharing
+   the same file).  Semantics (windows, budgets, success reset) are identical
+   to the in-memory store, which remains the default when the variable is
+   unset.
+
 Configuration via environment variables:
    OBSIGATE_LOGIN_MAX_ATTEMPTS     Max failures per IP (default: 10)
    OBSIGATE_ACCOUNT_MAX_ATTEMPTS   Max failures per account (default: 10)
    OBSIGATE_LOGIN_WINDOW_SECONDS   Lockout window in seconds (default: 900)
+    OBSIGATE_RATELIMIT_DB           SQLite file for shared/persistent counters (default: unset = memory)
 """

 import logging
 import os
+import sqlite3
+import threading
 import time
 from collections import defaultdict

@@ -37,6 +47,127 @@ _last_cleanup = time.time()
 CLEANUP_INTERVAL = 60  # seconds


+def _db_path() -> str | None:
+    """SQLite file for shared counters, or ``None`` for the in-memory store."""
+    path = os.environ.get("OBSIGATE_RATELIMIT_DB", "").strip()
+    return path or None
+
+
+def _db_connect(path: str) -> sqlite3.Connection:
+    """Open a short-lived connection (WAL + busy timeout for concurrent workers)."""
+    _db_ensure_schema(path)
+    conn = sqlite3.connect(path, timeout=10.0)
+    conn.execute("PRAGMA busy_timeout=10000")
+    return conn
+
+
+_schema_ready: set[str] = set()
+_schema_lock = threading.Lock()
+
+
+def _db_ensure_schema(path: str) -> None:
+    """Create the store schema once per file (DDL under a process-wide lock)."""
+    with _schema_lock:
+        if path in _schema_ready:
+            return
+        conn = sqlite3.connect(path, timeout=10.0)
+        try:
+            conn.execute("PRAGMA journal_mode=WAL")
+            conn.execute(
+                "CREATE TABLE IF NOT EXISTS attempts"
+                " (kind TEXT NOT NULL, key TEXT NOT NULL, ts REAL NOT NULL, success INTEGER NOT NULL)"
+            )
+            conn.execute(
+                "CREATE INDEX IF NOT EXISTS idx_attempts_kind_key_ts"
+                " ON attempts (kind, key, ts)"
+            )
+            conn.commit()
+        finally:
+            conn.close()
+        _schema_ready.add(path)
+
+
+def _db_write(fn, *args):
+    """Run a write op, retrying once on lock contention (concurrent workers)."""
+    try:
+        return fn(*args)
+    except sqlite3.OperationalError as e:
+        if "locked" not in str(e).lower():
+            raise
+        time.sleep(0.05)
+        return fn(*args)
+
+
+def _db_prune(conn: sqlite3.Connection, cutoff: float) -> None:
+    """Drop expired entries (best-effort cap on disk growth)."""
+    conn.execute("DELETE FROM attempts WHERE ts <= ?", (cutoff,))
+
+
+def _db_record(kind: str, key: str, success: bool) -> int:
+    """Record one attempt in SQLite; return the live failure count."""
+    path = _db_path()
+    assert path is not None
+    now = time.time()
+    cutoff = now - WINDOW_SECONDS
+
+    def _write() -> int:
+        with _db_connect(path) as conn:
+            _db_prune(conn, cutoff)
+            if success:
+                # Mirror the in-memory reset: replace history with one success.
+                conn.execute("DELETE FROM attempts WHERE kind = ? AND key = ?", (kind, key))
+            conn.execute(
+                "INSERT INTO attempts (kind, key, ts, success) VALUES (?, ?, ?, ?)",
+                (kind, key, now, int(success)),
+            )
+            conn.commit()
+            (failures,) = conn.execute(
+                "SELECT COUNT(*) FROM attempts WHERE kind = ? AND key = ? AND ts > ? AND success = 0",
+                (kind, key, cutoff),
+            ).fetchone()
+        return failures
+
+    return _db_write(_write)
+
+
+def _db_failures(kind: str, key: str) -> int:
+    """Live failure count in SQLite (expired entries never count)."""
+    path = _db_path()
+    assert path is not None
+    cutoff = time.time() - WINDOW_SECONDS
+    with _db_connect(path) as conn:
+        (failures,) = conn.execute(
+            "SELECT COUNT(*) FROM attempts WHERE kind = ? AND key = ? AND ts > ? AND success = 0",
+            (kind, key, cutoff),
+        ).fetchone()
+    return failures
+
+
+def _db_tracked(kind: str) -> int:
+    """Number of distinct keys ever seen for one budget (SQLite)."""
+    path = _db_path()
+    assert path is not None
+    with _db_connect(path) as conn:
+        (n,) = conn.execute(
+            "SELECT COUNT(DISTINCT key) FROM attempts WHERE kind = ?", (kind,)
+        ).fetchone()
+    return n
+
+
+def _db_limited_count(kind: str, max_attempts: int) -> int:
+    """Number of keys currently over budget (SQLite)."""
+    path = _db_path()
+    assert path is not None
+    cutoff = time.time() - WINDOW_SECONDS
+    with _db_connect(path) as conn:
+        rows = conn.execute(
+            "SELECT key, COUNT(*) FROM attempts"
+            " WHERE kind = ? AND ts > ? AND success = 0 GROUP BY key",
+            (kind, cutoff),
+        ).fetchall()
+    return sum(1 for _, n in rows if n >= max_attempts)
+
+
 def _prune(store: dict[str, list], cutoff: float) -> None:
    """Drop expired entries from one store in place."""
    expired = []
@@ -66,6 +197,12 @@ def record_failure(ip: str) -> tuple[int, int]:
    Returns:
        (current_failure_count, remaining_attempts)
    """
+    if _db_path() is not None:
+        failures = _db_record("ip", ip, False)
+        remaining = max(0, MAX_ATTEMPTS - failures)
+        if failures >= MAX_ATTEMPTS:
+            logger.warning(f"IP {ip} rate-limited after {failures} failed logins")
+        return failures, remaining
    _cleanup_expired()
    _ip_attempts[ip].append((time.time(), False))
    failures = sum(1 for _, success in _ip_attempts[ip] if not success)
@@ -77,12 +214,17 @@ def record_failure(ip: str) -> tuple[int, int]:

 def record_success(ip: str):
    """Clear rate limit state for an IP after successful login."""
+    if _db_path() is not None:
+        _db_record("ip", ip, True)
+        return
    _cleanup_expired()
    _ip_attempts[ip] = [(time.time(), True)]


 def is_rate_limited(ip: str) -> bool:
    """Check if an IP has exceeded the rate limit."""
+    if _db_path() is not None:
+        return _db_failures("ip", ip) >= MAX_ATTEMPTS
    _cleanup_expired()
    failures = sum(1 for _, success in _ip_attempts.get(ip, []) if not success)
    return failures >= MAX_ATTEMPTS
@@ -94,8 +236,14 @@ def record_account_failure(account: str) -> tuple[int, int]:
    Returns:
        (current_failure_count, remaining_attempts)
    """
-    _cleanup_expired()
    key = account.lower()
+    if _db_path() is not None:
+        failures = _db_record("account", key, False)
+        remaining = max(0, ACCOUNT_MAX_ATTEMPTS - failures)
+        if failures >= ACCOUNT_MAX_ATTEMPTS:
+            logger.warning(f"Account {account} rate-limited after {failures} failed attempts")
+        return failures, remaining
+    _cleanup_expired()
    _account_attempts[key].append((time.time(), False))
    failures = sum(1 for _, success in _account_attempts[key] if not success)
    remaining = max(0, ACCOUNT_MAX_ATTEMPTS - failures)
@@ -106,12 +254,17 @@ def record_account_failure(account: str) -> tuple[int, int]:

 def record_account_success(account: str):
    """Clear the per-account rate limit state after a successful login."""
+    if _db_path() is not None:
+        _db_record("account", account.lower(), True)
+        return
    _cleanup_expired()
    _account_attempts[account.lower()] = [(time.time(), True)]


 def is_account_rate_limited(account: str) -> bool:
    """Check if an account has exceeded the per-account rate limit."""
+    if _db_path() is not None:
+        return _db_failures("account", account.lower()) >= ACCOUNT_MAX_ATTEMPTS
    _cleanup_expired()
    failures = sum(
        1 for _, success in _account_attempts.get(account.lower(), []) if not success
@@ -121,6 +274,24 @@ def is_account_rate_limited(account: str) -> bool:

 def get_status(ip: str | None = None) -> dict:
    """Get rate limit status for an IP (for diagnostics)."""
+    if _db_path() is not None:
+        if ip:
+            failures = _db_failures("ip", ip)
+            return {
+                "ip": ip,
+                "failures": failures,
+                "max": MAX_ATTEMPTS,
+                "limited": failures >= MAX_ATTEMPTS,
+                "window_seconds": WINDOW_SECONDS,
+            }
+        return {
+            "tracked_ips": _db_tracked("ip"),
+            "tracked_accounts": _db_tracked("account"),
+            "max_attempts": MAX_ATTEMPTS,
+            "account_max_attempts": ACCOUNT_MAX_ATTEMPTS,
+            "window_seconds": WINDOW_SECONDS,
+            "limited_ips": _db_limited_count("ip", MAX_ATTEMPTS),
+        }
    _cleanup_expired()
    if ip:
        attempts = _ip_attempts.get(ip, [])
@@ -0,0 +1,210 @@
+"""Markdown rendering pipeline (ROADMAP #85, tranche 9).
+
+Helpers extraits de :mod:`backend.main` sans changement de comportement :
+slugification des headings, IDs d'ancrage, rendu mistune singleton,
+wikilinks, normalisation des sauts de ligne et pipeline complet
+:func:`_render_markdown` (rendu + sanitizer XSS BUG-021).
+
+Les noms gardent leur préfixe ``_`` d'origine pour un déplacement
+strictement verbatim (tests et routers pointent ici désormais).
+"""
+
+from __future__ import annotations
+
+import html as html_mod
+import re
+import unicodedata
+from pathlib import Path
+from typing import cast
+
+import mistune
+
+from backend.image_processor import preprocess_images
+from backend.indexer import find_file_in_index, get_vault_data
+from backend.secret_redactor import redact_file_content
+from backend.services.sanitizer import sanitize_html
+
+
+def _heading_slugify(text: str) -> str:
+    """Generate a URL-safe slug from heading text.
+
+    Matches the JavaScript slugify algorithm exactly using
+    Unicode-aware character classification:
+    1. Strip HTML tags (e.g. wikilink spans rendered inside headings)
+    2. Decode HTML entities (e.g. ``&amp;`` → ``&``)
+    3. Lowercase
+    4. NFD normalize + strip combining marks
+    5. Keep only Unicode letters, numbers, spaces, hyphens
+    6. Replace spaces with hyphens, collapse multiple hyphens
+
+    Args:
+        text: The heading text content (may contain inline HTML).
+
+    Returns:
+        A URL-safe slug string.
+    """
+    # Strip any inline HTML so it does not pollute the slug
+    text = re.sub(r"<[^>]+>", "", text)
+    # Decode HTML entities so &amp; becomes & before slugification
+    text = html_mod.unescape(text)
+    text = text.lower()
+    text = unicodedata.normalize("NFD", text)
+    text = "".join(ch for ch in text if not unicodedata.combining(ch))
+    # Unicode-aware: keep letters (L*), numbers (N*), spaces, and hyphens
+    cleaned = []
+    for ch in text:
+        cat = unicodedata.category(ch)
+        if cat.startswith('L') or cat.startswith('N') or ch in (' ', '-'):
+            cleaned.append(ch)
+    text = "".join(cleaned)
+    text = re.sub(r"\s+", "-", text)
+    text = re.sub(r"-+", "-", text)
+    result = text.strip("-")
+    return result if result else "heading"
+
+
+def _add_heading_ids(html: str) -> str:
+    """Post-process rendered HTML to add IDs to heading tags.
+
+    Adds an ``id`` attribute to every ``<h1>`` through ``<h6>`` tag
+    using a slug generated from the heading's text content.
+    Duplicate slugs get a ``-2``, ``-3``, etc. suffix.
+
+    Args:
+        html: Rendered HTML string.
+
+    Returns:
+        HTML with heading IDs injected.
+    """
+    used_ids: dict[str, int] = {}
+
+    def _replace_heading(match):
+        tag = match.group(1)
+        content = match.group(2)
+        slug = _heading_slugify(content)
+        count = used_ids.get(slug, 0)
+        used_ids[slug] = count + 1
+        if count > 0:
+            slug = f"{slug}-{count + 1}"
+        return f'<{tag} id="{slug}">{content}</{tag}>'
+
+    # Match h1-h6 tags with text content (no existing id attribute)
+    return re.sub(
+        r'<(h[1-6])>([^<]*(?:<(?!/?h[1-6])[^<]*)*)</h[1-6]>',
+        _replace_heading,
+        html,
+    )
+
+
+# Cached mistune renderer — avoids re-creating on every request
+_markdown_renderer = mistune.create_markdown(
+    escape=False,
+    plugins=["table", "strikethrough", "footnotes", "task_lists"],
+)
+
+
+def _convert_wikilinks(content: str, current_vault: str) -> str:
+    """Convert ``[[wikilinks]]`` and ``[[target|display]]`` to clickable HTML.
+
+    Supports:
+    - Internal file links: ``[[My Note]]`` / ``[[My Note|display]]``
+    - Same-document anchors: ``[[#Heading]]`` / ``[[#Heading|display]]``
+
+    Resolved file links get a ``data-vault`` / ``data-path`` attribute pair.
+    Anchor links target the slugified heading ID in the current document.
+    Unresolved links are rendered as ``<span class="wikilink-missing">``.
+
+    Args:
+        content: Markdown string potentially containing wikilinks.
+        current_vault: Active vault name for resolution priority.
+
+    Returns:
+        Markdown string with wikilinks replaced by HTML anchors.
+    """
+    def _replace(match):
+        target = match.group(1).strip()
+        display = match.group(2).strip() if match.group(2) else target
+
+        # Same-document anchor link: [[#Heading|display]]
+        if target.startswith("#"):
+            anchor_text = target[1:].strip()
+            anchor_slug = _heading_slugify(anchor_text)
+            link_display = display if display != target else anchor_text
+            return f'<a class="wikilink-anchor" href="#{anchor_slug}">{link_display}</a>'
+
+        found = find_file_in_index(target, current_vault)
+        if found:
+            return (
+                f'<a class="wikilink" href="#" '
+                f'data-vault="{found["vault"]}" '
+                f'data-path="{found["path"]}">{display}</a>'
+            )
+        return f'<span class="wikilink-missing">{display}</span>'
+
+    pattern = r'\[\[([^\]|]+)(?:\|([^\]]+))?\]\]'
+    return re.sub(pattern, _replace, content)
+
+
+def _normalize_line_breaks(text: str) -> str:
+    """Convert single newlines to hard breaks (matching Obsidian default behavior).
+
+    In standard Markdown, a single ``\\n`` is a "soft break" — it renders as a space,
+    not a visible line break.  Obsidian defaults to treating single newlines as hard
+    breaks (equivalent to ``<br>``).  This function pre-processes the Markdown source
+    so that mistune renders standalone lines on separate rows, while still honouring
+    blank lines as paragraph separators.
+
+    Fenced code blocks (`` ``` ``) are left untouched so their internal newlines are
+    preserved verbatim.
+    """
+    parts = re.split(r"(```[\s\S]*?```)", text)
+    for i, part in enumerate(parts):
+        if part.startswith("```"):
+            continue  # Protect fenced code blocks
+        # Single \n (not preceded or followed by another \n) → two spaces + \n
+        parts[i] = re.sub(r"(?<!\n)\n(?!\n)", "  \n", part)
+    return "".join(parts)
+
+
+def _render_markdown(raw_md: str, vault_name: str, current_file_path: Path | None = None) -> str:
+    """Render a markdown string to HTML with wikilink and image support.
+
+    Uses the cached singleton mistune renderer for performance.
+
+    Args:
+        raw_md: Raw markdown text (frontmatter already stripped).
+        vault_name: Current vault for wikilink resolution context.
+        current_file_path: Absolute path to the current markdown file.
+
+    Returns:
+        HTML string.
+    """
+    # Get vault data for image resolution
+    vault_data = get_vault_data(vault_name)
+    vault_root = Path(vault_data["path"]) if vault_data else None
+    attachments_path = vault_data.get("config", {}).get("attachmentsPath") if vault_data else None
+
+    # Redact secrets before rendering (P0 security)
+    raw_md = redact_file_content(raw_md, str(current_file_path) if current_file_path else "")
+
+    # Preprocess images first
+    if vault_root:
+        raw_md = preprocess_images(raw_md, vault_name, vault_root, current_file_path, attachments_path)
+
+    # Convert wikilinks
+    converted = _convert_wikilinks(raw_md, vault_name)
+
+    # Normalize line breaks to match Obsidian behavior (single \n → hard break)
+    converted = _normalize_line_breaks(converted)
+
+    # mistune 3.3 types `Markdown.__call__` as `str | list[...]` (les
+    # renderers HTML renvoient toujours `str` à l'exécution).
+    rendered = cast(str, _markdown_renderer(converted))
+
+    # Add heading IDs for TOC navigation
+    rendered = _add_heading_ids(rendered)
+
+    # Sanitize: raw HTML in vault content must never reach the DOM (BUG-021).
+    rendered = sanitize_html(rendered)
+
+    return rendered
@@ -1,9 +1,9 @@
-fastapi==0.110.3
-uvicorn==0.30.0
+fastapi==0.141.1
+uvicorn==0.54.0
 websockets>=12.0
 python-frontmatter==1.1.0
-mistune==3.0.2
-python-multipart==0.0.9
+mistune==3.3.3
+python-multipart==0.0.31
 aiofiles==23.2.1
 aiohttp>=3.9.0
 watchdog>=4.0.0
@@ -11,7 +11,7 @@ argon2-cffi>=23.1.0
 python-jose>=3.3.0
 sortedcontainers>=2.4.0
 snowballstemmer>=2.2.0
-weasyprint>=60.0
+weasyprint>=70.0
 httpx>=0.27.0
 pypdf>=4.0
 pyotp>=2.10.0
@@ -19,7 +19,7 @@ segno>=1.5.0
 webauthn==2.6.0
 psutil>=5.9
 pywebpush>=2.3.0
-mcp==1.9.4
+mcp==1.28.1
 sse-starlette==2.1.3
 openpyxl>=3.1
 python-docx>=1.1
@@ -0,0 +1,7 @@
+"""ObsiGate — routers FastAPI par domaine (ROADMAP #85).
+
+Découpage progressif du monolithe ``backend/main.py`` : chaque module de ce
+paquet expose un ``APIRouter`` monté par ``main.py``. Les handlers sont
+déplacés sans changement de comportement (mêmes chemins, mêmes modèles de
+réponse, mêmes dépendances d'authentification).
+"""
@@ -0,0 +1,412 @@
+"""Backup endpoints (ROADMAP #85, tranche 4).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/file/{vault}/backups|diff|restore``,
+``/api/backups*``), mêmes modèles de réponse, mêmes dépendances
+d'authentification. La logique métier vit déjà dans
+:mod:`backend.services.backups`.
+
+Adaptations strictement équivalentes :
+- ``_resolve_safe_path`` / ``_backup_file`` / ``_list_backup_files`` de
+  ``main`` n'étaient que des wrappers directs : appelés ici via
+  :mod:`backend.services.paths` et :mod:`backend.services.backups`.
+- ``RestoreRequest`` / ``RestoreResponse`` / ``DiffResponse`` ont déménagé
+  dans :mod:`backend.schemas`.
+- Le singleton SSE vit désormais dans :mod:`backend.sse` (partagé avec
+  ``main`` : les clients ``/api/events`` reçoivent les mêmes broadcasts).
+"""
+
+import logging
+import os
+import time
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+
+from fastapi import APIRouter, Body, Depends, HTTPException, Query
+
+from backend.auth.middleware import check_vault_access, require_auth
+from backend.indexer import get_vault_data, index, update_single_file
+from backend.schemas import (
+    BackupContentResponse,
+    BackupsAutoResponse,
+    BackupsCompressResponse,
+    BackupsDeletedResponse,
+    BackupsListResponse,
+    BackupsResponse,
+    DiffResponse,
+    RestoreRequest,
+    RestoreResponse,
+)
+from backend.services.backups import (
+    create_backup,
+)
+from backend.services.backups import (
+    diff_backup as service_diff_backup,
+)
+from backend.services.backups import (
+    list_backup_files as service_list_backup_files,
+)
+from backend.services.mutations import (
+    restore_backup as service_restore_backup,
+)
+from backend.services.paths import resolve_safe_path
+from backend.sse import sse_manager
+from backend.webhooks import dispatch_webhooks
+
+logger = logging.getLogger("obsigate")
+
+router = APIRouter(tags=["backups"])
+
+
+@router.get("/api/file/{vault_name}/backups", response_model=BackupsResponse)
+async def api_file_backups(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to file"),
+    current_user=Depends(require_auth),
+):
+    """List all available backups for a file.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative path of the file within the vault.
+
+    Returns:
+        BackupListResponse with backups sorted newest first.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"File not found: {path}")
+
+    try:
+        backups = service_list_backup_files(vault_name, path)
+    except Exception as e:
+        logger.error(f"Error listing backups for {vault_name}/{path}: {type(e).__name__}: {e}", exc_info=True)
+        raise HTTPException(status_code=500, detail=f"Erreur lors de la lecture des backups: {e!s}")
+
+    return {"vault": vault_name, "path": path, "backups": backups}
+
+
+@router.get("/api/file/{vault_name}/diff", response_model=DiffResponse)
+async def api_file_diff(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to file"),
+    version: int = Query(..., description="Timestamp of the backup version (left/old side)"),
+    compare_with: int | None = Query(default=None, description="Timestamp of another backup (right/new side). If omitted, compares with the current file."),
+    current_user=Depends(require_auth),
+):
+    """Generate a unified diff between a backup version and another version or the current file.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative path of the file within the vault.
+        version: Timestamp of the backup to use as the old/left side.
+        compare_with: Optional timestamp of another backup as the new/right side.
+                       If omitted, the current file on disk is used.
+
+    Returns:
+        DiffResponse containing the unified diff string.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    return service_diff_backup(vault_name, path, version, compare_with)
+
+
+@router.post("/api/file/{vault_name}/restore", response_model=RestoreResponse)
+async def api_file_restore(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to file"),
+    body: RestoreRequest = ..., # type: ignore
+    current_user=Depends(require_auth),
+):
+    """Restore a file from a backup version.
+
+    The current file is backed up before being overwritten (so the operation is reversible).
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative path of the file within the vault.
+        body: RestoreRequest with the backup version timestamp.
+
+    Returns:
+        RestoreResponse confirming the restore.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_restore_backup(vault_name, path, body.version)
+    current_backed_up = result["current_backed_up"]
+
+    # Update index
+    await update_single_file(vault_name, path)
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("file_restored", {
+        "vault": vault_name,
+        "path": path,
+        "restored_from": body.version,
+        "current_backed_up": current_backed_up,
+    })
+    await dispatch_webhooks("file_restored", {"vault": vault_name, "path": path, "restored_from": body.version})
+
+    return {
+        "success": True,
+        "vault": vault_name,
+        "path": path,
+        "restored_from": body.version,
+        "current_backed_up": current_backed_up,
+    }
+
+
+@router.get("/api/backups", response_model=BackupsListResponse)
+async def api_backups_list(
+    vault: str | None = Query(None, description="Filter by vault name"),
+    current_user=Depends(require_auth),
+):
+    """List all backups across vaults, grouped by file."""
+    result: list[dict[str, Any]] = []
+    try:
+      for vault_name in index:
+        if vault and vault_name != vault:
+            continue
+        if not check_vault_access(vault_name, current_user):
+            continue
+        vd = get_vault_data(vault_name)
+        if not vd:
+            continue
+        vault_root = Path(vd["path"])
+        backup_root = Path(os.environ.get("OBSIGATE_BACKUP_DIR", ".obsigate-backup"))
+        if not backup_root.is_absolute():
+            backup_root = vault_root / backup_root
+        vault_backup_dir = backup_root / vault_name
+        if not vault_backup_dir.exists():
+            continue
+        for fpath in vault_backup_dir.rglob("*.bak"):
+            if not fpath.is_file():
+                continue
+            st = fpath.stat()
+            fsize = st.st_size
+            ts_part = fpath.name.rsplit(".", 2)
+            if len(ts_part) < 3 or not ts_part[-2].isdigit():
+                continue
+            ts = int(ts_part[-2])
+            rel_dir = str(fpath.parent.relative_to(vault_backup_dir)).replace("\\", "/")
+            rel_file = rel_dir + "/" + ts_part[0] if rel_dir != "." else ts_part[0]
+            result.append({
+                "vault": vault_name,
+                "file": rel_file,
+                "backup_file": fpath.name,
+                "timestamp": ts,
+                "datetime": datetime.fromtimestamp(ts, tz=timezone.utc).isoformat(),
+                "size": fsize,
+                "full_path": str(fpath),
+            })
+
+      result.sort(key=lambda x: x["timestamp"], reverse=True)
+      total_size = sum(r["size"] for r in result)
+      return {"backups": result, "total": len(result), "total_size_bytes": total_size}
+    except Exception as e:
+        logger.error(f"Error listing backups: {type(e).__name__}: {e}", exc_info=True)
+        raise HTTPException(status_code=500, detail=f"Erreur listing backups: {e!s}")
+
+
+@router.post("/api/backups/delete", response_model=BackupsDeletedResponse)
+async def api_backups_delete(
+    body: dict = Body(...),
+    current_user=Depends(require_auth),
+):
+    """Delete one or more backup files."""
+    paths = body.get("paths", [])
+    if not paths:
+        raise HTTPException(status_code=400, detail="No backup paths provided")
+
+    deleted = 0
+    for p in paths:
+        try:
+            fpath = Path(p)
+            # Security: ensure path is within a backup directory
+            if ".obsigate-backup" not in str(fpath):
+                continue
+            if fpath.exists() and fpath.is_file():
+                fpath.unlink()
+                deleted += 1
+        except Exception as e:
+            logger.warning(f"Failed to delete backup {p}: {e}")
+
+    return {"deleted": deleted}
+
+
+@router.post("/api/backups/purge", response_model=BackupsDeletedResponse)
+async def api_backups_purge(
+    body: dict = Body(...),
+    current_user=Depends(require_auth),
+):
+    """Purge all backups for a specific file or entire vault."""
+    vault_name = body.get("vault")
+    file_path = body.get("file")  # optional
+
+    if not vault_name:
+        raise HTTPException(status_code=400, detail="Vault name required")
+
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail="Access denied")
+
+    vd = get_vault_data(vault_name)
+    if not vd:
+        raise HTTPException(status_code=404, detail="Vault not found")
+
+    vault_root = Path(vd["path"])
+    backup_root = Path(os.environ.get("OBSIGATE_BACKUP_DIR", ".obsigate-backup"))
+    if not backup_root.is_absolute():
+        backup_root = vault_root / backup_root
+
+    if file_path:
+        # Delete backups for specific file
+        backup_dir = backup_root / vault_name / Path(file_path).parent
+        if backup_dir.exists():
+            fname = Path(file_path).name
+            deleted = 0
+            for f in backup_dir.iterdir():
+                if f.is_file() and f.name.startswith(fname + ".") and f.name.endswith(".bak"):
+                    f.unlink()
+                    deleted += 1
+            return {"deleted": deleted}
+        return {"deleted": 0}
+    else:
+        # Delete all backups for vault
+        vault_backup_dir = backup_root / vault_name
+        if vault_backup_dir.exists():
+            deleted = 0
+            for f in vault_backup_dir.rglob("*.bak"):
+                if f.is_file():
+                    f.unlink()
+                    deleted += 1
+            return {"deleted": deleted}
+        return {"deleted": 0}
+
+
+
+@router.get("/api/backups/content", response_model=BackupContentResponse)
+async def api_backups_content(
+    path: str = Query(..., description="Full path to backup file"),
+    current_user=Depends(require_auth),
+):
+    """Return the content of a specific backup file."""
+    try:
+        fpath = Path(path)
+        if ".obsigate-backup" not in str(fpath):
+            raise HTTPException(status_code=403, detail="Access denied")
+        if not fpath.exists() or not fpath.is_file():
+            raise HTTPException(status_code=404, detail="Backup not found")
+        content = fpath.read_text(encoding="utf-8", errors="replace")
+        # Truncate large files to 100KB
+        if len(content) > 102400:
+            content = content[:102400] + "\n\n... (tronque a 100 Ko)"
+        return {"content": content, "name": fpath.name, "size": len(content)}
+    except HTTPException:
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+@router.post("/api/backups/compress", response_model=BackupsCompressResponse)
+async def api_backups_compress(
+    body: dict = Body(...),
+    current_user=Depends(require_auth),
+):
+    """Compress backups older than N days. Body: {older_than_days: 30, dry_run: false}"""
+    import gzip as gz_mod
+    older_than = body.get("older_than_days", 30)
+    dry_run = body.get("dry_run", False)
+    cutoff = time.time() - (older_than * 86400)
+    compressed = 0
+    saved_bytes = 0
+
+    for vault_name in index:
+        if not check_vault_access(vault_name, current_user):
+            continue
+        vd = get_vault_data(vault_name)
+        if not vd:
+            continue
+        vault_root = Path(vd["path"])
+        backup_root = Path(os.environ.get("OBSIGATE_BACKUP_DIR", ".obsigate-backup"))
+        if not backup_root.is_absolute():
+            backup_root = vault_root / backup_root
+        vault_dir = backup_root / vault_name
+        if not vault_dir.exists():
+            continue
+        for fpath in vault_dir.rglob("*.bak"):
+            if not fpath.is_file():
+                continue
+            if fpath.name.endswith(".bak.gz"):
+                continue
+            mtime = fpath.stat().st_mtime
+            if mtime > cutoff:
+                continue
+            if not dry_run:
+                try:
+                    gz_path = fpath.with_suffix(fpath.suffix + ".gz")
+                    data = fpath.read_bytes()
+                    with gz_mod.open(str(gz_path), "wb", compresslevel=6) as gzf:
+                        gzf.write(data)
+                    orig_size = len(data)
+                    gz_size = gz_path.stat().st_size
+                    if gz_size < orig_size:
+                        fpath.unlink()
+                        saved_bytes += (orig_size - gz_size)
+                    else:
+                        gz_path.unlink()  # compression didn't help
+                    compressed += 1
+                except Exception as e:
+                    logger.warning(f"Failed to compress {fpath}: {e}")
+            else:
+                compressed += 1
+
+    return {"compressed": compressed, "saved_bytes": saved_bytes, "dry_run": dry_run}
+
+
+@router.post("/api/backups/auto", response_model=BackupsAutoResponse)
+async def api_backups_auto(
+    body: dict = Body(...),
+    current_user=Depends(require_auth),
+):
+    """Create backups for files modified since a given time. Body: {since_hours: 24}"""
+    since_hours = body.get("since_hours", 24)
+    cutoff = time.time() - (since_hours * 3600)
+    backed_up = 0
+
+    for vault_name in index:
+        if not check_vault_access(vault_name, current_user):
+            continue
+        vd = get_vault_data(vault_name)
+        if not vd:
+            continue
+        vault_root = Path(vd["path"])
+        for fpath in vault_root.rglob("*"):
+            if not fpath.is_file():
+                continue
+            if fpath.name.startswith('.'):
+                continue
+            if any(p.startswith('.') or p in {'.obsidian', '.trash', '.git', '.obsigate-backup', '__pycache__', 'node_modules'} for p in fpath.relative_to(vault_root).parts):
+                continue
+            mtime = fpath.stat().st_mtime
+            if mtime < cutoff:
+                continue
+            try:
+                rel = str(fpath.relative_to(vault_root)).replace("\\", "/")
+                create_backup(fpath, vault_name, rel)
+                backed_up += 1
+            except Exception as e:
+                logger.warning(f"Auto-backup failed for {rel}: {e}")
+
+    return {"backed_up": backed_up, "since_hours": since_hours}
@@ -0,0 +1,531 @@
+"""Configuration, AI keys, diagnostics & dashboard endpoints (ROADMAP #85, tranche 7).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/config*``, ``/api/diagnostics``,
+``/api/dashboard``), mêmes modèles de réponse, mêmes dépendances
+d'authentification.
+
+Adaptations strictement équivalentes :
+- ``_load_config`` / ``_save_config`` / ``_DEFAULT_CONFIG`` /
+  ``_CONFIG_PATH`` / ``_BASE_DIR`` ont déménagé ici : ``main`` les
+  réimporte pour son lifespan (pas de cycle : ce module ne dépend pas de
+  ``main``).
+- ``AI_KEYS_FILE`` / ``_write_ai_keys`` / ``_FALLBACK_MODELS`` ont déménagé
+  ici (``AI_KEYS_FILE`` garde son chemin relatif ``data/api_keys.json``,
+  résolu depuis le même CWD au runtime).
+"""
+
+import json as _json
+import logging
+import os
+import urllib.request
+from pathlib import Path
+
+from fastapi import APIRouter, Body, Depends, HTTPException, Query
+
+from backend.ai import PROVIDERS, _read_ai_keys, get_ai_key
+from backend.auth.middleware import require_admin, require_auth
+from backend.indexer import index
+from backend.media_types import IMAGE_EXTENSIONS
+from backend.schemas import (
+    AIKeyDeleteResponse,
+    AIKeysResponse,
+    AIModelsResponse,
+    AITestResponse,
+    AppConfigResponse,
+    DashboardResponse,
+    DiagnosticsResponse,
+    StatusResponse,
+)
+from backend.search_executor import get_search_executor
+from backend.tools.secrets import (
+    TOOL_KEY_NAMES as _TOOL_KEY_NAMES,
+)
+from backend.tools.secrets import (
+    delete_tool_key as _delete_tool_key,
+)
+from backend.tools.secrets import (
+    get_tool_key as _get_tool_key,
+)
+from backend.tools.secrets import (
+    mask_value as _mask_tool_value,
+)
+from backend.tools.secrets import (
+    set_tool_key as _set_tool_key,
+)
+
+logger = logging.getLogger("obsigate")
+
+router = APIRouter(tags=["System"])
+
+_BASE_DIR = Path(__file__).resolve().parent.parent.parent
+_CONFIG_PATH = _BASE_DIR / "data" / "config.json"
+
+_DEFAULT_CONFIG = {
+    "search_workers": 2,
+    "debounce_ms": 300,
+    "results_per_page": 50,
+    "min_query_length": 2,
+    "search_timeout_ms": 30000,
+    "max_content_size": 100000,
+    "snippet_context_chars": 120,
+    "max_snippet_highlights": 5,
+    "title_boost": 3.0,
+    "path_boost": 1.5,
+    "watcher_enabled": True,
+    "watcher_use_polling": False,
+    "watcher_polling_interval": 5.0,
+    "watcher_debounce": 2.0,
+    "tag_boost": 2.0,
+    "prefix_max_expansions": 50,
+    "recent_files_limit": 20,
+    "max_backups_per_file": 10,
+    "ai_default_provider": "deepseek",
+    "ai_default_models": {},
+}
+
+
+def _load_config() -> dict:
+    """Load config from disk, merging with defaults."""
+    config = dict(_DEFAULT_CONFIG)
+    if _CONFIG_PATH.exists():
+        try:
+            stored = _json.loads(_CONFIG_PATH.read_text(encoding="utf-8"))
+            config.update(stored)
+        except Exception as e:
+            logger.warning(f"Failed to read config.json: {e}")
+    return config
+
+
+def _save_config(config: dict) -> None:
+    """Persist config to disk."""
+    try:
+        _CONFIG_PATH.write_text(
+            _json.dumps(config, indent=2, ensure_ascii=False),
+            encoding="utf-8",
+        )
+    except Exception as e:
+        logger.error(f"Failed to write config.json: {e}")
+        raise HTTPException(status_code=500, detail=f"Failed to save config: {e}")
+
+
+AI_KEYS_FILE = Path("data/api_keys.json")
+
+def _write_ai_keys(data: dict):
+    AI_KEYS_FILE.parent.mkdir(parents=True, exist_ok=True)
+    tmp = AI_KEYS_FILE.with_suffix(".tmp")
+    tmp.write_text(_json.dumps(data, indent=2), encoding="utf-8")
+    tmp.replace(AI_KEYS_FILE)
+
+@router.get("/api/config", response_model=AppConfigResponse)
+async def api_get_config(current_user=Depends(require_auth)):
+    """Return current configuration with defaults for missing keys."""
+    return _load_config()
+
+
+@router.post("/api/config", response_model=AppConfigResponse)
+async def api_set_config(body: dict = Body(...), current_user=Depends(require_admin)):
+    """Update configuration. Only known keys are accepted.
+
+    Keys matching ``_DEFAULT_CONFIG`` are validated and persisted.
+    Unknown keys are silently ignored.
+    Returns the full merged config after update.
+    """
+    current = _load_config()
+    updated_keys = []
+    for key, value in body.items():
+        if key in _DEFAULT_CONFIG:
+            expected_type = type(_DEFAULT_CONFIG[key])
+            if isinstance(value, expected_type) or (expected_type is float and isinstance(value, (int, float))):
+                current[key] = value
+                updated_keys.append(key)
+            else:
+                raise HTTPException(
+                    status_code=400,
+                    detail=f"Invalid type for '{key}': expected {expected_type.__name__}, got {type(value).__name__}",
+                )
+    _save_config(current)
+    if any(k.startswith("ai_") for k in updated_keys):
+        try:
+            from backend.ai import reload_ai_config
+            reload_ai_config()
+        except Exception as e:
+            logger.warning(f"Failed to reload AI config: {e}")
+    logger.info(f"Config updated: {updated_keys}")
+    return current
+
+
+@router.get("/api/config/ai-keys", response_model=AIKeysResponse)
+async def api_get_ai_keys(current_user=Depends(require_admin)):
+    """Return stored AI keys (values masked)."""
+    keys = _read_ai_keys()
+    masked = {}
+    for k in ["DEEPSEEK_API_KEY", "OPENROUTER_API_KEY", "GEMINI_API_KEY", "NVIDIA_API_KEY", "QWENCLOUD_API_KEY", "XIAOMI_API_KEY", "MISTRAL_API_KEY"]:
+        val = keys.get(k, "") or os.environ.get(k, "")
+        if val:
+            masked[k] = val[:4] + "..." + val[-4:] if len(val) > 8 else "***"
+        else:
+            masked[k] = ""
+    return masked
+
+@router.post("/api/config/ai-keys", response_model=StatusResponse)
+async def api_set_ai_keys(body: dict = Body(...), current_user=Depends(require_admin)):
+    """Save AI keys. Pass {"DEEPSEEK_API_KEY":"sk-...","OPENROUTER_API_KEY":"...","GEMINI_API_KEY":"..."}"""
+    keys = _read_ai_keys()
+    for k in ["DEEPSEEK_API_KEY", "OPENROUTER_API_KEY", "GEMINI_API_KEY", "NVIDIA_API_KEY", "QWENCLOUD_API_KEY", "XIAOMI_API_KEY", "MISTRAL_API_KEY"]:
+        if body.get(k):
+            keys[k] = body[k]
+    _write_ai_keys(keys)
+    logger.info("AI keys updated")
+    return {"status": "ok"}
+
+
+@router.delete("/api/config/ai-keys/{provider_env}", response_model=AIKeyDeleteResponse)
+async def api_delete_ai_key(provider_env: str, current_user=Depends(require_admin)):
+    """Delete a specific AI provider key from storage."""
+    allowed = {"DEEPSEEK_API_KEY", "OPENROUTER_API_KEY", "GEMINI_API_KEY",
+               "NVIDIA_API_KEY", "QWENCLOUD_API_KEY", "XIAOMI_API_KEY", "MISTRAL_API_KEY"}
+    key_name = provider_env.upper()
+    if key_name not in allowed:
+        raise HTTPException(status_code=400, detail=f"Clé inconnue: {provider_env}")
+    keys = _read_ai_keys()
+    if key_name in keys:
+        del keys[key_name]
+        _write_ai_keys(keys)
+    # Also clear from env at runtime so get_ai_key() no longer finds it
+    os.environ.pop(key_name, None)
+    logger.info(f"AI key deleted: {key_name}")
+    return {"status": "deleted", "key": key_name}
+
+
+@router.get("/api/config/tool-keys", response_model=AIKeysResponse)
+async def api_get_tool_keys(current_user=Depends(require_admin)):
+    """Return tool/connected-source configuration (tokens masked, URLs clear)."""
+    masked = {}
+    for name in _TOOL_KEY_NAMES:
+        masked[name] = _mask_tool_value(name, _get_tool_key(name))
+    return masked
+
+
+@router.post("/api/config/tool-keys", response_model=StatusResponse)
+async def api_set_tool_keys(body: dict = Body(...), current_user=Depends(require_admin)):
+    """Save tool/connected-source keys.
+
+    Only whitelisted names (``backend.tools.secrets.TOOL_KEY_NAMES``) are
+    accepted: Tavily/Brave/SerpAPI/Exa API keys, Gitea URL + token, GitHub
+    token. Empty values delete the stored entry.
+    """
+    updated = []
+    for name, value in body.items():
+        if name not in _TOOL_KEY_NAMES:
+            raise HTTPException(status_code=400, detail=f"Clé inconnue: {name}")
+        if value is not None and not isinstance(value, str):
+            raise HTTPException(status_code=400, detail=f"Type invalide pour {name}")
+        _set_tool_key(name, value or "")
+        updated.append(name)
+    logger.info(f"Tool keys updated: {updated}")
+    return {"status": "ok"}
+
+
+@router.delete("/api/config/tool-keys/{name}", response_model=AIKeyDeleteResponse)
+async def api_delete_tool_key(name: str, current_user=Depends(require_admin)):
+    """Delete a stored tool key (the environment fallback still applies)."""
+    key_name = name.upper()
+    try:
+        existed = _delete_tool_key(key_name)
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+    logger.info(f"Tool key deleted: {key_name} (existed={existed})")
+    return {"status": "deleted", "key": key_name}
+
+
+@router.post("/api/config/ai-keys/test", response_model=AITestResponse)
+async def api_test_ai_keys(current_user=Depends(require_admin)):
+    """Test which AI providers are configured.
+
+    Each provider has a dedicated (URL, header-name) test pair.
+    - Most OpenAI-compatible APIs use `Authorization: Bearer KEY`
+    - Xiaomi MiMo uses `api-key: KEY`
+    - Gemini uses a query-string key
+    """
+    results = {}
+    for key_name, label, test_url_tmpl, header_name in [
+        # OpenAI-compatible — Authorization: Bearer
+        ("DEEPSEEK_API_KEY",  "deepseek",   "https://api.deepseek.com/v1/models",                   "Authorization"),
+        ("OPENROUTER_API_KEY","openrouter", "https://openrouter.ai/api/v1/models",                   "Authorization"),
+        ("NVIDIA_API_KEY",    "nvidia",     "https://integrate.api.nvidia.com/v1/models",           "Authorization"),
+        ("QWENCLOUD_API_KEY", "qwencloud",  "https://dashscope.aliyuncs.com/compatible-mode/v1/models", "Authorization"),
+        ("MISTRAL_API_KEY",   "mistral",    "https://api.mistral.ai/v1/models",                     "Authorization"),
+        # Xiaomi MiMo — dedicated api-key header (NOT Authorization: Bearer)
+        ("XIAOMI_API_KEY",    "xiaomi",     "https://api.xiaomimimo.com/v1/models",                 "api-key"),
+        # Gemini — key in query string
+        ("GEMINI_API_KEY",    "gemini",     "https://generativelanguage.googleapis.com/v1beta/models?key={key}", None),
+    ]:
+        key = get_ai_key(key_name)
+        if not key:
+            results[label] = "non configuré"
+            continue
+        try:
+            url = test_url_tmpl.replace("{key}", key) if "{key}" in test_url_tmpl else test_url_tmpl
+            if header_name:
+                req = urllib.request.Request(url, headers={header_name: key})
+            else:
+                req = urllib.request.Request(url)
+            urllib.request.urlopen(req, timeout=5)
+            results[label] = "ok"
+        except Exception as e:
+            # Truncate the error to keep the response small.
+            results[label] = "erreur: " + str(e)[:80]
+    return results
+
+
+@router.get("/api/config/ai-models", response_model=AIModelsResponse)
+async def api_list_ai_models(provider: str = Query(...), current_user=Depends(require_admin)):
+    """List available models for a given AI provider.
+
+    Strategy:
+    1. Try the provider's public models endpoint (OpenAI-compatible /v1/models or Gemini).
+    2. If the network call fails (timeout, 4xx, 5xx, DNS, etc.), fall back to a
+       curated static list of known-good models for that provider.
+    3. Always return a non-empty list when the provider is known, so the UI
+       dropdown is never empty.
+    """
+    provider = provider.lower()
+
+    from backend.model_capabilities import get_capabilities_for_models
+    from backend.provider_capabilities import remember_declared_capabilities
+
+    all_providers = ("deepseek", "openrouter", "gemini", "nvidia", "qwencloud", "xiaomi", "mistral")
+    if provider not in all_providers:
+        return {"models": [], "error": f"Unknown provider: {provider}", "source": "validation"}
+
+    key_name = f"{provider.upper()}_API_KEY"
+    key = get_ai_key(key_name)
+    if not key:
+        # No key configured — return curated fallback list so the UI can
+        # still show what WOULD be available once a key is set.
+        fallback = _FALLBACK_MODELS.get(provider, [])
+        return {"models": fallback, "source": "fallback",
+                "capabilities": get_capabilities_for_models(provider, fallback),
+                "note": "API key not configured — showing default model list"}
+
+    # Build URL
+    if provider == "gemini":
+        url = f"https://generativelanguage.googleapis.com/v1beta/models?key={key}"
+    elif provider == "deepseek":
+        url = "https://api.deepseek.com/v1/models"
+    elif provider == "openrouter":
+        url = "https://openrouter.ai/api/v1/models"
+    elif provider == "nvidia":
+        url = "https://integrate.api.nvidia.com/v1/models"
+    elif provider == "qwencloud":
+        url = "https://dashscope.aliyuncs.com/compatible-mode/v1/models"
+    elif provider == "xiaomi":
+        # Xiaomi MiMo — dedicated api-key header (NOT Authorization: Bearer).
+        # Endpoint: https://api.xiaomimimo.com/v1/models
+        url = "https://api.xiaomimimo.com/v1/models"
+        models = []  # parsed below with the custom header
+    elif provider == "mistral":
+        url = "https://api.mistral.ai/v1/models"
+
+    try:
+        if provider == "gemini":
+            req = urllib.request.Request(url)
+        elif provider == "xiaomi":
+            # Xiaomi MiMo uses a dedicated api-key header.
+            req = urllib.request.Request(url, headers={"api-key": key})
+        else:
+            req = urllib.request.Request(url, headers={"Authorization": "Bearer " + key})
+
+        with urllib.request.urlopen(req, timeout=10) as resp:
+            data = _json.loads(resp.read().decode())
+
+        if provider == "gemini":
+            models = [m.get("name", "") for m in data.get("models", []) if m.get("name")]
+            # Gemini returns names like "models/gemini-1.5-flash" — strip prefix
+            models = [m.replace("models/", "") for m in models]
+        else:
+            models = [m.get("id", "") for m in data.get("data", []) if m.get("id")]
+
+        # Cache the capabilities the provider declares for these models
+        # (BUG-044) — get_capabilities_for_models() below then returns the
+        # provider's own truth for the flags it declares, the curated table
+        # for the rest. Providers that declare nothing are left untouched.
+        remember_declared_capabilities(provider, data)
+
+        if models:
+            # Prepend the configured default if not already present
+            default = PROVIDERS.get(provider, {}).get("model")
+            if default and default not in models:
+                models = [default] + models
+            return {"models": models, "source": "live", "count": len(models),
+                    "capabilities": get_capabilities_for_models(provider, models)}
+        # Empty list from API — fall through to fallback
+        raise ValueError("empty model list from provider API")
+    except Exception as e:
+        # Network error, auth error, parsing error — use curated fallback
+        fallback = _FALLBACK_MODELS.get(provider, [])
+        return {"models": fallback, "source": "fallback", "error": str(e)[:200],
+                "capabilities": get_capabilities_for_models(provider, fallback),
+                "note": "Could not reach provider API — showing default model list"}
+
+
+# ── Curated fallback model lists ──────────────────────────────────────────
+# Used when the provider API is unreachable or returns empty.
+# Keep these short and focused on models known to work with the
+# OpenAI-compatible chat completions interface (or Gemini's generateContent).
+_FALLBACK_MODELS: dict[str, list[str]] = {
+    "deepseek": [
+        "deepseek-chat",
+        "deepseek-reasoner",
+    ],
+    "openrouter": [
+        "openai/gpt-4o-mini",
+        "openai/gpt-4o",
+        "anthropic/claude-3.5-sonnet",
+        "anthropic/claude-3-haiku",
+        "google/gemini-2.0-flash-exp:free",
+        "meta-llama/llama-3.1-70b-instruct",
+        "meta-llama/llama-3.1-8b-instruct:free",
+        "mistralai/mistral-large-latest",
+    ],
+    "gemini": [
+        "gemini-2.0-flash",
+        "gemini-2.0-flash-exp",
+        "gemini-1.5-pro",
+        "gemini-1.5-flash",
+        "gemini-1.5-flash-8b",
+    ],
+    "nvidia": [
+        "meta/llama-3.1-405b-instruct",
+        "meta/llama-3.1-70b-instruct",
+        "meta/llama-3.1-8b-instruct",
+        "mistralai/mistral-large",
+        "google/gemma-2-27b-it",
+        "nvidia/llama-3.1-nemotron-70b-instruct",
+    ],
+    "qwencloud": [
+        "qwen-max",
+        "qwen-plus",
+        "qwen-turbo",
+        "qwen-long",
+        "qwen-vl-max",
+        "qwen-vl-plus",
+    ],
+    "xiaomi": [
+        # Xiaomi MiMo models — the public /v1/models endpoint requires the
+        # `api-key` custom header (NOT Authorization: Bearer), so the live
+        # call often fails with 401 even with the right key. We ship a
+        # known-good list as fallback. See https://mimo.mi.com/docs/
+        "mimo-v2.5-pro",
+        "mimo-v2.5",
+        "mimo-v2.5-asr",
+        "mimo-v2.5-tts",
+        "mimo-v2.5-tts-voiceclone",
+        "mimo-v2.5-tts-voicedesign",
+    ],
+    "mistral": [
+        "mistral-large-latest",
+        "mistral-medium-latest",
+        "mistral-small-latest",
+        "open-mistral-7b",
+        "open-mixtral-8x7b",
+        "codestral-latest",
+    ],
+}
+
+
+@router.get("/api/diagnostics", response_model=DiagnosticsResponse)
+async def api_diagnostics(current_user=Depends(require_admin)):
+    """Return index statistics and system diagnostics.
+
+    Includes document counts, token counts, memory estimates,
+    and inverted index status.
+    """
+    import sys
+
+    from backend.search import get_inverted_index
+
+    inv = get_inverted_index()
+
+    # Per-vault stats
+    vault_stats = {}
+    total_files = 0
+    total_tags = 0
+    # Snapshot both dicts first: the indexer mutates them from background
+    # threads, and iterating a live dict raises "dictionary changed size".
+    for vname, vdata in list(index.items()):
+        file_count = len(vdata.get("files", []))
+        tag_count = len(vdata.get("tags", {}))
+        vault_stats[vname] = {"file_count": file_count, "tag_count": tag_count}
+        total_files += file_count
+        total_tags += tag_count
+
+    # Memory estimate for inverted index
+    word_index = inv.word_index.copy()
+    word_index_entries = sum(len(docs) for docs in word_index.values())
+    mem_estimate_mb = round(
+        (sys.getsizeof(inv.word_index) + word_index_entries * 80
+         + len(inv.doc_info) * 200
+         + len(inv._sorted_tokens) * 60) / (1024 * 1024), 2
+    )
+
+    return {
+        "index": {
+            "total_files": total_files,
+            "total_tags": total_tags,
+            "vaults": vault_stats,
+        },
+        "inverted_index": {
+            "unique_tokens": len(word_index),
+            "total_postings": word_index_entries,
+            "documents": inv.doc_count,
+            "sorted_tokens": len(inv._sorted_tokens),
+            "is_ready": inv.is_ready(),
+            "memory_estimate_mb": mem_estimate_mb,
+        },
+        "config": _load_config(),
+        "search_executor": {
+            "active": get_search_executor() is not None,
+            "max_workers": get_search_executor()._max_workers if get_search_executor() else 0,
+        },
+    }
+
+
+@router.get("/api/dashboard", response_model=DashboardResponse)
+async def api_dashboard(current_user=Depends(require_auth)):
+    """Aggregated dashboard statistics across all accessible vaults."""
+    user_vaults = current_user.get("_token_vaults") or current_user.get("vaults", [])
+    vault_stats = []
+    total_files = 0
+    total_tags = set()
+    total_size = 0
+    total_images = 0
+    for vname, vdata in index.items():
+        if "*" not in user_vaults and vname not in user_vaults:
+            continue
+        files = vdata.get("files", [])
+        fc = len(files)
+        total_files += fc
+        vtags = set()
+        vsize = 0
+        vimages = 0
+        for f in files:
+            vtags.update(f.get("tags", []))
+            vsize += f.get("size", 0)
+            if (f.get("extension") or "").lower() in IMAGE_EXTENSIONS:
+                vimages += 1
+        total_tags.update(vtags)
+        total_size += vsize
+        total_images += vimages
+        vault_stats.append({
+            "name": vname, "file_count": fc, "tag_count": len(vtags),
+            "total_size_bytes": vsize, "image_count": vimages,
+        })
+    return {
+        "vaults": vault_stats,
+        "total_files": total_files,
+        "total_tags": len(total_tags),
+        "total_size_bytes": total_size,
+        "total_images": total_images,
+    }
@@ -0,0 +1,73 @@
+"""Syncthing conflict endpoints (ROADMAP #85, tranche 8).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/conflicts*``), mêmes modèles de
+réponse, mêmes dépendances d'authentification.
+
+Adaptations strictement équivalentes :
+- ``_resolve_safe_path`` / ``_backup_file`` → :mod:`backend.services.paths`
+  et :mod:`backend.services.backups` (pass-through).
+"""
+
+import logging
+import shutil
+from pathlib import Path
+
+from fastapi import APIRouter, Body, Depends, HTTPException
+
+from backend.audit import log_file_delete
+from backend.auth.middleware import check_vault_access, require_auth
+from backend.indexer import get_conflicts, get_vault_data, remove_single_file
+from backend.schemas import ConflictResolveResponse, ConflictsResponse
+from backend.services.backups import create_backup
+from backend.services.paths import resolve_safe_path
+from backend.sse import sse_manager
+
+logger = logging.getLogger("obsigate")
+
+router = APIRouter(tags=["conflicts"])
+
+
+@router.get("/api/conflicts", response_model=ConflictsResponse)
+async def api_conflicts(current_user=Depends(require_auth)):
+    """List sync-conflict files across accessible vaults."""
+    user_vaults = current_user.get("_token_vaults") or current_user.get("vaults", [])
+    all_conflicts = get_conflicts()
+    if "*" not in user_vaults:
+        all_conflicts = [c for c in all_conflicts if c["vault"] in user_vaults]
+    return {"conflicts": all_conflicts, "total": len(all_conflicts)}
+
+
+@router.post("/api/conflicts/resolve", response_model=ConflictResolveResponse)
+async def api_conflict_resolve(body: dict = Body(...), current_user=Depends(require_auth)):
+    """Resolve a conflict: keep_local (delete conflict file) or keep_conflict (replace original)."""
+    vault_name = body.get("vault")
+    conflict_path = body.get("conflict_path")
+    original_path = body.get("original_path")
+    action = body.get("action")  # "keep_local" or "keep_conflict"
+    # mypy: narrow down from dict values
+    assert isinstance(vault_name, str), "'vault' is required and must be a string"
+    assert isinstance(conflict_path, str), "'conflict_path' is required and must be a string"
+    assert isinstance(original_path, str), "'original_path' is required and must be a string"
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(403, f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(404, "Vault not found")
+    vault_root = Path(vault_data["path"])
+    conf_file = resolve_safe_path(vault_root, conflict_path)
+    orig_file = resolve_safe_path(vault_root, original_path)
+    if not conf_file.exists():
+        raise HTTPException(404, "Conflict file not found")
+    try:
+        if action == "keep_conflict":
+            create_backup(orig_file, vault_name, original_path)
+            shutil.copy2(conf_file, orig_file)
+            logger.info(f"Conflict resolved (keep_conflict): {conflict_path} → {original_path}")
+        conf_file.unlink()
+        await remove_single_file(vault_name, conflict_path)
+        log_file_delete(current_user["username"], vault_name, conflict_path)
+        await sse_manager.broadcast("file_deleted", {"vault": vault_name, "path": conflict_path})
+        return {"status": "resolved", "action": action}
+    except Exception as e:
+        raise HTTPException(500, f"Error resolving conflict: {e!s}")
@@ -0,0 +1,569 @@
+"""Media, PDF, export & vault-settings endpoints (ROADMAP #85, tranche 6c).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/file/*/pdf*``, ``/api/export/*``,
+``/api/guide/download``, ``/api/image/*``, ``/api/media*``,
+``/api/attachments/*``, ``/api/vaults/*/settings``, ``/api/vault/*/files``,
+``/api/vaults/settings/all``), mêmes modèles de réponse, mêmes dépendances
+d'authentification.
+
+Adaptations strictement équivalentes :
+- ``_resolve_safe_path`` → :mod:`backend.services.paths` (pass-through).
+- ``_render_markdown`` vient de :mod:`backend.render` (#85 T9, sans cycle
+  d'import).
+- ``_resolve_export_target`` / ``_safe_export_name`` (export uniquement)
+  sont définis ici ; ``stream_file_with_range`` vit dans
+  :mod:`backend.routers.helpers` (partagé).
+"""
+
+import asyncio
+import logging
+from pathlib import Path
+
+from fastapi import APIRouter, Body, Depends, HTTPException, Query, Request
+from fastapi.responses import FileResponse, Response
+
+from backend.attachment_indexer import get_attachment_stats, rescan_vault_attachments
+from backend.auth.middleware import check_vault_access, require_admin, require_auth
+from backend.export import ExportError, export_epub, export_html, export_md_bundle
+from backend.history import record_open
+from backend.indexer import get_vault_data, index, parse_markdown_file
+from backend.media_thumbs import generate_thumbnail, is_decodable
+from backend.media_types import is_audio, is_image, is_video, media_mime_type
+from backend.render import _render_markdown
+from backend.routers.helpers import media_max_inline_bytes, stream_file_with_range
+from backend.schemas import (
+    AllVaultSettingsResponse,
+    AttachmentRescanResponse,
+    AttachmentStatsResponse,
+    PdfInfoResponse,
+    VaultFilesResponse,
+    VaultSettingsResponse,
+)
+from backend.secret_redactor import redact_file_content
+from backend.services.paths import resolve_safe_path
+from backend.services.vaults import list_all_files
+from backend.vault_settings import get_vault_setting, update_vault_setting
+
+logger = logging.getLogger("obsigate")
+
+# Lazy import: WeasyPrint PDF export (requires GTK, may not be available everywhere)
+try:
+    from backend.pdf_export import build_pdf_html, generate_pdf
+except Exception:  # pragma: no cover - WeasyPrint/GTK missing
+    generate_pdf = None  # type: ignore[assignment]
+    build_pdf_html = None  # type: ignore[assignment]
+
+    logging.getLogger("obsigate").warning("PDF export unavailable (WeasyPrint/GTK not found)")
+
+router = APIRouter()  # pas de tags : assignation par chemin via openapi_docs.tag_for_path (comme avant)
+
+
+def _resolve_export_target(vault_name: str, path: str, current_user: dict) -> tuple[Path, Path]:
+    """Resolve a vault + relative path into (vault_root, absolute file path).
+
+    Enforces auth (vault access) and path traversal protection.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+    vault_root = Path(vault_data["path"])
+    target = resolve_safe_path(vault_root, path)
+    return vault_root, target
+
+
+def _safe_export_name(name: str) -> str:
+    """ASCII-safe, filename-safe download name (falls back to 'document')."""
+    cleaned = "".join(c for c in name if c.isascii() and (c.isalnum() or c in " _-.")).strip()
+    return cleaned or "document"
+
+
+@router.get(
+    "/api/file/{vault_name}/pdf",
+    response_class=Response,
+    responses={200: {"content": {"application/pdf": {}}, "description": "PDF document"}},
+)
+async def api_file_pdf(vault_name: str, path: str = Query(..., description="Relative path to file"), current_user=Depends(require_auth)):
+    """Download a markdown file as PDF."""
+    if generate_pdf is None:
+        raise HTTPException(501, "PDF export unavailable (WeasyPrint/GTK not available)")
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(403, f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(404, f"Vault '{vault_name}' not found")
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+    if not file_path.exists():
+        raise HTTPException(404, f"File not found: {path}")
+    try:
+        raw = file_path.read_text(encoding="utf-8", errors="replace")
+    except Exception:
+        raise HTTPException(500, "Cannot read file")
+    record_open(current_user.get("username"), vault_name, path)
+    raw = redact_file_content(raw, str(file_path))
+    post = parse_markdown_file(raw)
+    html = _render_markdown(post.content, vault_name, file_path)
+    title = post.metadata.get("title", file_path.stem)
+    pdf_html = build_pdf_html(html, str(title))
+    pdf_bytes = generate_pdf(pdf_html, str(title))
+    safe_name = "".join(c for c in str(title) if c.isascii() and (c.isalnum() or c in " _-.")).strip() or "document"
+    return Response(content=pdf_bytes, media_type="application/pdf", headers={"Content-Disposition": f'attachment; filename="{safe_name}.pdf"'})
+
+
+@router.get(
+    "/api/export/html",
+    response_class=Response,
+    responses={200: {"content": {"text/html": {}}, "description": "Standalone HTML file"}},
+)
+async def api_export_html(
+    vault: str = Query(..., description="Vault name"),
+    path: str = Query(..., description="Relative path to file"),
+    current_user=Depends(require_auth),
+):
+    """Export a markdown note as a standalone HTML file."""
+    try:
+        vault_root, target = _resolve_export_target(vault, path, current_user)
+        html_bytes = export_html(vault_root, target)
+    except ExportError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+    record_open(current_user.get("username"), vault, path)
+    safe_name = _safe_export_name(target.stem)
+    return Response(
+        content=html_bytes,
+        media_type="text/html; charset=utf-8",
+        headers={"Content-Disposition": f'attachment; filename="{safe_name}.html"'},
+    )
+
+
+@router.get(
+    "/api/export/md-bundle",
+    response_class=Response,
+    responses={200: {"content": {"application/zip": {}}, "description": "Markdown ZIP bundle"}},
+)
+async def api_export_md_bundle(
+    vault: str = Query(..., description="Vault name"),
+    path: str = Query(..., description="Relative path to directory or file"),
+    current_user=Depends(require_auth),
+):
+    """Export a directory (or single file) of markdown as a ZIP bundle."""
+    try:
+        vault_root, target = _resolve_export_target(vault, path, current_user)
+        zip_bytes = export_md_bundle(vault_root, target)
+    except ExportError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+    safe_name = _safe_export_name(target.name)
+    return Response(
+        content=zip_bytes,
+        media_type="application/zip",
+        headers={"Content-Disposition": f'attachment; filename="{safe_name}.zip"'},
+    )
+
+
+@router.get(
+    "/api/export/epub",
+    response_class=Response,
+    responses={200: {"content": {"application/epub+zip": {}}, "description": "ePub document"}},
+)
+async def api_export_epub(
+    vault: str = Query(..., description="Vault name"),
+    path: str = Query(..., description="Relative path to file"),
+    current_user=Depends(require_auth),
+):
+    """Export a markdown note as an ePub document."""
+    try:
+        vault_root, target = _resolve_export_target(vault, path, current_user)
+        epub_bytes = export_epub(vault_root, target)
+    except ExportError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+    record_open(current_user.get("username"), vault, path)
+    safe_name = _safe_export_name(target.stem)
+    return Response(
+        content=epub_bytes,
+        media_type="application/epub+zip",
+        headers={"Content-Disposition": f'attachment; filename="{safe_name}.epub"'},
+    )
+
+
+@router.get(
+    "/api/guide/download",
+    response_class=Response,
+    responses={200: {"content": {"application/pdf": {}, "text/markdown": {}}}},
+)
+async def api_guide_download(
+    format: str = Query("md", description="Download format: 'md' or 'pdf'"),
+    lang: str = Query("fr", description="Guide language: 'fr' or 'en'"),
+    current_user=Depends(require_auth),
+):
+    """Download the in-app user guide as Markdown or PDF (#105).
+
+    The document is generated from the live help modal in index.html resolved
+    through the locale files, so it always mirrors exactly what the user sees.
+    """
+    from backend.guide_export import get_guide_document
+
+    if format not in ("md", "pdf"):
+        raise HTTPException(status_code=400, detail="format doit être 'md' ou 'pdf'")
+    try:
+        payload, media, fname = get_guide_document(format, lang)
+    except Exception as e:  # weasyprint/reportlab unavailable
+        logger.exception("guide export failed")
+        raise HTTPException(status_code=500, detail=f"Export impossible: {e}") from e
+    return Response(
+        content=payload,
+        media_type=media,
+        headers={"Content-Disposition": f'attachment; filename="{fname}"'},
+    )
+
+
+@router.get("/api/file/{vault_name}/pdf/stream", response_class=FileResponse)
+async def api_pdf_stream(
+    request: Request,
+    vault_name: str,
+    path: str = Query(...),
+    current_user=Depends(require_auth),
+):
+    """Stream a PDF file with Content-Type: application/pdf for inline browser viewing.
+
+    Supports HTTP Range requests (206 Partial Content) so browsers can
+    progressively render large PDFs in the native viewer.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"File not found: {path}")
+    if file_path.suffix.lower() != ".pdf":
+        raise HTTPException(status_code=400, detail="Not a PDF file")
+
+    return stream_file_with_range(file_path, request, "application/pdf")
+
+
+@router.get("/api/file/{vault_name}/pdf/info", response_model=PdfInfoResponse)
+async def api_pdf_info(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to PDF file"),
+    current_user=Depends(require_auth),
+):
+    """Return PDF metadata (pages, title, author, size) without the document content.
+
+    Lets the UI display file info before loading a heavy PDF into the viewer.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"File not found: {path}")
+    if file_path.suffix.lower() != ".pdf":
+        raise HTTPException(status_code=400, detail="Not a PDF file")
+
+    from backend.pdf_reader import extract_pdf_metadata
+    meta = extract_pdf_metadata(file_path)
+    stat = file_path.stat()
+    return {
+        "vault": vault_name,
+        "path": path,
+        "pages": meta.get("pages", 0),
+        "title": meta.get("title") or file_path.name,
+        "author": meta.get("author", ""),
+        "size_bytes": stat.st_size,
+    }
+
+
+@router.get(
+    "/api/image/{vault_name}",
+    response_class=Response,
+    responses={200: {"content": {"application/octet-stream": {}}, "description": "Image bytes"}},
+)
+async def api_image(vault_name: str, path: str = Query(..., description="Relative path to image"), current_user=Depends(require_auth)):
+    """Serve an image file with proper MIME type.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative file path within the vault.
+
+    Returns:
+        Image file with appropriate content-type header.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"Image not found: {path}")
+
+    mime_type = media_mime_type(str(file_path))
+
+    # #108-B3 — a standalone SVG opened in a tab executes its embedded JS
+    # (same-origin XSS). ``sandbox`` forces a unique opaque origin with no
+    # script execution; inside an <img> tag the header is irrelevant.
+    headers = {"X-Content-Type-Options": "nosniff"}
+    if file_path.suffix.lower() == ".svg":
+        headers["Content-Security-Policy"] = "sandbox"
+
+    try:
+        # Read and return the image file
+        content = file_path.read_bytes()
+        return Response(content=content, media_type=mime_type, headers=headers)
+    except PermissionError:
+        raise HTTPException(status_code=403, detail="Permission denied")
+    except Exception as e:
+        logger.error(f"Error serving image {vault_name}/{path}: {e}")
+        raise HTTPException(status_code=500, detail=f"Error serving image: {e!s}")
+
+
+@router.get("/api/media/{vault_name}", response_class=FileResponse)
+async def api_media_stream(
+    request: Request,
+    vault_name: str,
+    path: str = Query(..., description="Relative path to audio/video file"),
+    current_user=Depends(require_auth),
+):
+    """Stream an audio/video file with HTTP Range support (roadmap #109-A2).
+
+    Serves the bytes with the correct MIME type and honours ``Range`` requests
+    (``206 Partial Content`` + ``Content-Range``/``Accept-Ranges``), which is
+    what enables scrubbing in ``<audio>``/``<video>`` and is required by Safari
+    for MP4. Files above ``OBSIGATE_MEDIA_MAX_INLINE_MB`` (default 500 MB) are
+    refused with ``413`` — the viewer falls back to the download button.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"Media not found: {path}")
+
+    ext = file_path.suffix.lower()
+    if not (is_audio(ext) or is_video(ext)):
+        raise HTTPException(status_code=400, detail="Not an audio/video file")
+
+    if file_path.stat().st_size > media_max_inline_bytes():
+        raise HTTPException(status_code=413, detail="Media too large for inline streaming")
+
+    return stream_file_with_range(file_path, request, media_mime_type(str(file_path)))
+
+
+@router.get("/api/media/{vault_name}/thumb", response_class=FileResponse)
+async def api_media_thumb(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to image"),
+    size: int = Query(256, ge=32, le=1024, description="Max thumbnail edge in pixels"),
+    current_user=Depends(require_auth),
+):
+    """Serve a cached WebP thumbnail of an image (roadmap #108-C).
+
+    SVG (and any format Pillow cannot decode) falls back to the original
+    bytes. Generation runs in a thread and is capped at 2 s; on timeout or
+    failure the original is served so the UI never breaks.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"Image not found: {path}")
+    if not is_image(file_path.suffix.lower()):
+        raise HTTPException(status_code=400, detail="Not an image file")
+
+    mime_type = media_mime_type(str(file_path))
+    if not is_decodable(file_path):
+        # SVG: never let a standalone navigation execute embedded JS (#108-B3).
+        svg_headers = {"X-Content-Type-Options": "nosniff"}
+        if file_path.suffix.lower() == ".svg":
+            svg_headers["Content-Security-Policy"] = "sandbox"
+        return FileResponse(str(file_path), media_type=mime_type, headers=svg_headers)
+
+    loop = asyncio.get_running_loop()
+    thumb: Path | None = None
+    try:
+        thumb = await asyncio.wait_for(
+            loop.run_in_executor(None, generate_thumbnail, file_path, size),
+            timeout=2.0,
+        )
+    except Exception:
+        thumb = None
+
+    if thumb is not None and thumb.exists():
+        return FileResponse(str(thumb), media_type="image/webp")
+    return FileResponse(str(file_path), media_type=mime_type)
+
+
+@router.post("/api/attachments/rescan/{vault_name}", response_model=AttachmentRescanResponse)
+async def api_rescan_attachments(vault_name: str, current_user=Depends(require_admin)):
+    """Rescan attachments for a specific vault.
+
+    Args:
+        vault_name: Name of the vault to rescan.
+
+    Returns:
+        Dict with status and attachment count.
+    """
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    vault_path = vault_data["path"]
+    count = await rescan_vault_attachments(vault_name, vault_path)
+
+    logger.info(f"Rescanned attachments for vault '{vault_name}': {count} attachments")
+    return {"status": "ok", "vault": vault_name, "attachment_count": count}
+
+
+@router.get("/api/attachments/stats", response_model=AttachmentStatsResponse)
+async def api_attachment_stats(vault: str | None = Query(None, description="Vault filter"), current_user=Depends(require_auth)):
+    """Get attachment statistics for vaults.
+
+    Args:
+        vault: Optional vault name to filter stats.
+
+    Returns:
+        Dict with vault names as keys and attachment counts as values.
+    """
+    stats = get_attachment_stats(vault)
+    return {"vaults": stats}
+
+
+@router.get("/api/vaults/{vault_name}/settings", response_model=VaultSettingsResponse)
+async def api_get_vault_settings(vault_name: str, current_user=Depends(require_auth)):
+    """Get UI display settings for a specific vault.
+
+    Args:
+        vault_name: Name of the vault.
+
+    Returns:
+        Dict with vault settings including hideHiddenFiles.
+    """
+    if vault_name not in index:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    # Get persisted settings
+    persisted = get_vault_setting(vault_name) or {}
+
+    # Default settings
+    settings = {
+        "hideHiddenFiles": False,
+    }
+    settings.update(persisted)
+
+    return settings
+
+
+@router.post("/api/vaults/{vault_name}/settings", response_model=VaultSettingsResponse)
+async def api_update_vault_settings(vault_name: str, body: dict = Body(...), current_user=Depends(require_admin)):
+    """Update UI display settings for a specific vault.
+
+    Args:
+        vault_name: Name of the vault.
+        body: Dict with settings to update (hideHiddenFiles).
+
+    Returns:
+        Updated settings dict.
+    """
+    if vault_name not in index:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    # Validate settings
+    settings_to_update = {}
+
+    if "hideHiddenFiles" in body:
+        if not isinstance(body["hideHiddenFiles"], bool):
+            raise HTTPException(status_code=400, detail="hideHiddenFiles must be a boolean")
+        settings_to_update["hideHiddenFiles"] = body["hideHiddenFiles"]
+
+    # Update persisted settings
+    try:
+        updated = update_vault_setting(vault_name, settings_to_update)
+    except PermissionError as e:
+        logger.error(f"Permission error saving settings for vault '{vault_name}': {e}")
+        raise HTTPException(
+            status_code=500,
+            detail="Permission denied: Cannot write to settings file. Check /app/data permissions."
+        )
+    except Exception as e:
+        logger.error(f"Error saving settings for vault '{vault_name}': {e}")
+        raise HTTPException(
+            status_code=500,
+            detail=f"Failed to save settings: {e!s}"
+        )
+
+    logger.info(f"Updated settings for vault '{vault_name}': {settings_to_update}")
+
+    return updated
+
+
+@router.get("/api/vault/{vault_name}/files", response_model=VaultFilesResponse)
+async def api_vault_recent_files(
+    vault_name: str,
+    dir: str = Query("", description="Directory path within the vault (empty = root)"),
+    limit: int = Query(200, description="Maximum number of files to return"),
+    recursive: bool = Query(True, description="If true, list files recursively from directory and all subdirectories"),
+    current_user=Depends(require_auth),
+):
+    """List files in a vault directory sorted by modification time (newest first).
+
+    Returns file metadata suitable for a vault home page display.
+    Unlike /api/browse, this endpoint sorts by mtime and returns
+    additional metadata (size, modified time, extension).
+
+    When recursive=True (default), lists files from the directory
+    AND all its subdirectories, with a ``rel_dir`` field indicating
+    the subdirectory path relative to the requested directory.
+
+    Args:
+        vault_name: Name of the vault.
+        dir: Relative directory path within the vault (empty for root).
+        limit: Maximum files to return (default 200).
+        recursive: If true, recursively list files in subdirectories (default true).
+
+    Returns:
+        JSON with vault, directory, count, recursive flag, and list of file entries.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    return list_all_files(vault_name, dir=dir, limit=limit, recursive=recursive)
+
+
+@router.get("/api/vaults/settings/all", response_model=AllVaultSettingsResponse)
+async def api_get_all_vault_settings(current_user=Depends(require_auth)):
+    """Get UI display settings for all vaults.
+
+    Returns:
+        Dict mapping vault names to their settings.
+    """
+    all_settings = {}
+
+    for vault_name in index:
+        persisted = get_vault_setting(vault_name) or {}
+
+        settings = {
+            "hideHiddenFiles": False,
+        }
+        settings.update(persisted)
+        all_settings[vault_name] = settings
+
+    return all_settings
@@ -0,0 +1,593 @@
+"""File browsing & reading endpoints (ROADMAP #85, tranche 6a).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/browse/*``, ``/api/file/*`` en
+lecture), mêmes modèles de réponse (déménagés dans
+:mod:`backend.schemas`), mêmes dépendances d'authentification.
+
+Adaptations strictement équivalentes :
+- ``_resolve_safe_path`` → :mod:`backend.services.paths` (pass-through).
+- ``_render_markdown`` vient de :mod:`backend.render` (#85 T9, sans cycle
+  d'import).
+- ``_content_disposition`` / ``_media_max_inline_bytes`` / ``EXT_TO_LANG``
+  ont déménagé : helpers partagés dans :mod:`backend.routers.helpers`
+  (``EXT_TO_LANG`` n'était utilisé que par la vue fichier).
+"""
+
+import html as html_mod
+import logging
+from pathlib import Path
+from urllib.parse import quote
+
+from fastapi import APIRouter, Depends, HTTPException, Query
+from fastapi.responses import FileResponse
+
+from backend.auth.middleware import check_vault_access, require_auth
+from backend.history import record_open
+from backend.indexer import (
+    _extract_tags,
+    get_backlinks,
+    get_vault_data,
+    parse_markdown_file,
+)
+from backend.media_types import is_audio, is_image, is_video, media_mime_type
+from backend.render import _render_markdown
+from backend.routers.helpers import media_max_inline_bytes
+from backend.schemas import (
+    BacklinksResponse,
+    BrowseResponse,
+    FileContentResponse,
+    FileRawResponse,
+    XlsxSheetWindowResponse,
+)
+from backend.services.files import read_raw_file
+from backend.services.paths import resolve_safe_path
+from backend.services.vaults import browse_directory
+
+logger = logging.getLogger("obsigate")
+
+# Map file extensions to highlight.js language hints
+EXT_TO_LANG = {
+    ".py": "python", ".js": "javascript", ".ts": "typescript",
+    ".jsx": "jsx", ".tsx": "tsx", ".sh": "bash", ".bash": "bash",
+    ".zsh": "bash", ".fish": "fish", ".bat": "batch", ".cmd": "batch",
+    ".ps1": "powershell", ".json": "json", ".yaml": "yaml", ".yml": "yaml",
+    ".toml": "toml", ".xml": "xml", ".csv": "plaintext",
+    ".cfg": "ini", ".ini": "ini", ".conf": "ini", ".env": "bash",
+    ".html": "html", ".css": "css", ".scss": "scss", ".less": "less",
+    ".java": "java", ".c": "c", ".cpp": "cpp", ".h": "c", ".hpp": "cpp",
+    ".cs": "csharp", ".go": "go", ".rs": "rust", ".rb": "ruby",
+    ".php": "php", ".sql": "sql", ".r": "r", ".swift": "swift",
+    ".kt": "kotlin", ".txt": "plaintext", ".log": "plaintext",
+    ".lua": "lua", ".pl": "perl", ".pm": "perl", ".ex": "elixir", ".exs": "elixir",
+    ".dart": "dart", ".tf": "haskell", ".gradle": "groovy", ".groovy": "groovy",
+    ".graphql": "graphql", ".gql": "graphql", ".prisma": "sql", ".proto": "c",
+    ".vb": "basic", ".asm": "x86asm", ".s": "armasm",
+    ".vue": "xml", ".svelte": "xml", ".astro": "xml",
+    ".properties": "ini", ".service": "ini", ".hosts": "ini",
+    ".ksh": "bash", ".dockerfile": "dockerfile",
+    ".makefile": "makefile", ".cmake": "cmake",
+}
+
+router = APIRouter(tags=["files"])
+
+
+@router.get("/api/browse/{vault_name}", response_model=BrowseResponse)
+async def api_browse(vault_name: str, path: str = "", current_user=Depends(require_auth)):
+    """Browse directories and files in a vault at a given path level.
+
+    Returns sorted entries (directories first, then files) with metadata.
+    Hidden files/directories (starting with ``"."`` ) are excluded.
+
+    Args:
+        vault_name: Name of the vault to browse.
+        path: Relative directory path within the vault (empty = root).
+
+    Returns:
+        ``BrowseResponse`` with vault name, path, and item list.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    return browse_directory(vault_name, path)
+
+
+@router.get("/api/file/{vault_name}/raw", response_model=FileRawResponse)
+async def api_file_raw(vault_name: str, path: str = Query(..., description="Relative path to file"), current_user=Depends(require_auth)):
+    """Return raw file content as plain text.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative file path within the vault.
+
+    Returns:
+        ``FileRawResponse`` with vault, path, and raw text content.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    return read_raw_file(vault_name, path)
+
+
+@router.get("/api/file/{vault_name}/download", response_class=FileResponse)
+async def api_file_download(vault_name: str, path: str = Query(..., description="Relative path to file"), current_user=Depends(require_auth)):
+    """Download a file as an attachment.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative file path within the vault.
+
+    Returns:
+        ``FileResponse`` with ``application/octet-stream`` content-type.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"File not found: {path}")
+
+    # Record history
+    record_open(current_user.get("username"), vault_name, path)
+
+    return FileResponse(
+        path=str(file_path),
+        filename=file_path.name,
+        media_type="application/octet-stream",
+    )
+
+
+@router.get("/api/file/{vault_name}/backlinks", response_model=BacklinksResponse)
+async def api_file_backlinks(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to file"),
+    current_user=Depends(require_auth),
+):
+    """Get backlinks (files linking to this file via wikilinks).
+
+    Returns a list of files that contain `[[wikilinks]]` pointing
+    to the requested file, across all accessible vaults.
+
+    Args:
+        vault_name: Name of the vault containing the target file.
+        path: Relative path of the target file within the vault.
+
+    Returns:
+        ``{"vault": str, "path": str, "backlinks": [...]}``
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    user_vaults = current_user.get("_token_vaults") or current_user.get("vaults", [])
+    backlinks = get_backlinks(vault_name, path)
+
+    # Filter by user-accessible vaults
+    if "*" not in user_vaults:
+        backlinks = [b for b in backlinks if b["vault"] in user_vaults]
+
+    return {
+        "vault": vault_name,
+        "path": path,
+        "backlinks": backlinks,
+        "total": len(backlinks),
+    }
+
+
+@router.get(
+    "/api/file/{vault_name}/xlsx/sheet", response_model=XlsxSheetWindowResponse
+)
+def api_file_xlsx_sheet(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to the .xlsx file"),
+    sheet: str = Query(..., description="Sheet name (as shown in the viewer tab)"),
+    offset: int = Query(0, ge=0, description="0-based index of the first row to return"),
+    limit: int = Query(
+        200, ge=1, le=1000, description="Rows to return (server-capped)"
+    ),
+    current_user=Depends(require_auth),
+):
+    """Return a window of rows of one sheet of an .xlsx workbook (#153 A9).
+
+    Backs the viewer's lazy loading: instead of every sheet in a single JSON
+    payload, the client asks for the block it is about to display. The row
+    numbers and the ``data-cell`` references are the real A1 coordinates of the
+    sheet, so a window behaves like the full render (editing a cell in it
+    targets the right cell).
+
+    The response also carries ``total_rows``/``total_cols`` and the ``truncated``
+    flag, so the client can say what is hidden behind the 500x40 render caps
+    instead of silently hiding it.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative path of the .xlsx file within the vault.
+        sheet: Sheet name; **404** if the workbook has no such sheet.
+        offset: 0-based index of the first row to return.
+        limit: Rows to return, capped server-side at 1000.
+
+    Returns:
+        ``XlsxSheetWindowResponse`` with the rendered ``html`` of the window.
+
+    Raises:
+        HTTPException: 403 (vault access), 404 (vault, file or sheet unknown),
+            415 (not an .xlsx file), 500 (unreadable workbook).
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    file_path = resolve_safe_path(Path(vault_data["path"]), path)
+    if not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"File not found: {path}")
+    if file_path.suffix.lower() != ".xlsx":
+        raise HTTPException(status_code=415, detail="Le fichier n'est pas un classeur .xlsx")
+
+    # Import tardif : openpyxl n'est chargé que si un .xlsx est réellement demandé.
+    from backend.xlsx_reader import read_sheet_window
+
+    try:
+        window = read_sheet_window(file_path, sheet, offset=offset, limit=limit)
+    except Exception as e:
+        logger.error(f"XLSX sheet read error for {path}: {e}")
+        raise HTTPException(status_code=500, detail=f"Error reading XLSX: {e!s}")
+    if window is None:
+        raise HTTPException(status_code=404, detail=f"Feuille introuvable: {sheet}")
+
+    return {"vault": vault_name, "path": path, **window}
+
+
+@router.get("/api/file/{vault_name}", response_model=FileContentResponse)
+async def api_file(vault_name: str, path: str = Query(..., description="Relative path to file"), current_user=Depends(require_auth)):
+    """Return rendered HTML and metadata for a file.
+
+    Markdown files are parsed for frontmatter, rendered with wikilink
+    support, and returned with extracted tags. Other supported file
+    types are syntax-highlighted as code blocks.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative file path within the vault.
+
+    Returns:
+        ``FileContentResponse`` with HTML, metadata, and tags.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    vault_data = get_vault_data(vault_name)
+    if not vault_data:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, path)
+
+    if not file_path.exists() or not file_path.is_file():
+        raise HTTPException(status_code=404, detail=f"File not found: {path}")
+
+    # Record history
+    record_open(current_user.get("username"), vault_name, path, title=file_path.name)
+
+    ext = file_path.suffix.lower()
+
+    # === PDF: special handling before read_text (binary file) ===
+    if ext == ".pdf":
+        try:
+            from backend.pdf_reader import extract_pdf_metadata, extract_pdf_text, extract_pdf_toc
+            pdf_text = extract_pdf_text(file_path, max_chars=100000)
+            pdf_meta = extract_pdf_metadata(file_path)
+            pdf_toc = extract_pdf_toc(file_path)
+            size = file_path.stat().st_size
+            return {
+                "vault": vault_name,
+                "path": path,
+                "title": pdf_meta.get("title") or file_path.name,
+                "tags": [],
+                "frontmatter": {},
+                "html": f"<div class='pdf-viewer'><p>PDF — {pdf_meta.get('pages', '?')} pages</p><pre>{pdf_text[:5000]}</pre></div>",
+                "raw_length": size,
+                "extension": ext,
+                "is_markdown": False,
+                "is_pdf": True,
+                "unsupported": False,
+                "pdf_metadata": pdf_meta,
+                "pdf_toc": pdf_toc,
+                "size_bytes": size,
+            }
+        except Exception as e:
+            logger.error(f"PDF read error for {path}: {e}")
+            raise HTTPException(status_code=500, detail=f"Error reading PDF: {e!s}")
+
+    # === Excel .xlsx: render sheets as HTML tables (binary, before read_text) ===
+    if ext == ".xlsx":
+        try:
+            from backend.xlsx_reader import inspect_workbook, render_sheets
+
+            sheets = render_sheets(file_path)
+            size = file_path.stat().st_size
+            return {
+                "vault": vault_name,
+                "path": path,
+                "title": file_path.name,
+                "tags": [],
+                "frontmatter": {},
+                "html": sheets[0]["html"] if sheets else "",
+                "raw_length": size,
+                "extension": ext,
+                "is_markdown": False,
+                "is_xlsx": True,
+                "xlsx_sheets": sheets,
+                # #153 A1 — parts a save would drop; the viewer warns and asks
+                # for an explicit confirmation before forcing the write.
+                "xlsx_lossy_features": inspect_workbook(file_path),
+                "unsupported": False,
+                "size_bytes": size,
+            }
+        except Exception as e:
+            logger.error(f"XLSX read error for {path}: {e}")
+            raise HTTPException(status_code=500, detail=f"Error reading XLSX: {e!s}")
+
+    # === Images: return as viewable image ===
+    if is_image(ext):
+        size = file_path.stat().st_size
+        mime = media_mime_type(str(file_path))
+        # #108-B1 — the raw endpoint returns JSON (FileRawResponse), so the
+        # standalone <img> must point to /api/image, which serves the bytes
+        # with the right MIME type. Paths are URL-encoded (accents, spaces).
+        img_url = f"/api/image/{quote(vault_name, safe='')}?path={quote(path, safe='')}"
+        html = (
+            f'<div class="image-viewer">'
+            f'<img src="{img_url}" '
+            f'alt="{html_mod.escape(file_path.name, quote=True)}" '
+            f'style="max-width:100%;max-height:80vh;object-fit:contain" />'
+            f'</div>'
+        )
+        return {
+            "vault": vault_name,
+            "path": path,
+            "title": file_path.name,
+            "tags": [],
+            "frontmatter": {},
+            "html": html,
+            "raw_length": size,
+            "extension": ext,
+            "is_markdown": False,
+            "is_image": True,
+            "image_mime": mime,
+            "size_bytes": size,
+        }
+
+    # === Audio / Video: HTML5 players streamed from /api/media (roadmap #109) ===
+    if is_audio(ext) or is_video(ext):
+        size = file_path.stat().st_size
+        mime = media_mime_type(str(file_path))
+        media_kind = "audio" if is_audio(ext) else "video"
+
+        # #109-A3 — beyond the inline limit the viewer falls back to download
+        # (a single uvicorn worker must not be pinned by multi-GB media).
+        if size > media_max_inline_bytes():
+            return {
+                "vault": vault_name,
+                "path": path,
+                "title": file_path.name,
+                "tags": [],
+                "frontmatter": {},
+                "html": "",
+                "raw_length": size,
+                "extension": ext,
+                "is_markdown": False,
+                "unsupported": True,
+                "media_too_large": True,
+                "size_bytes": size,
+            }
+
+        # #109-A2 — byte-range endpoint: enables scrub and is required by Safari.
+        stream_url = f"/api/media/{quote(vault_name, safe='')}?path={quote(path, safe='')}"
+        if media_kind == "audio":
+            html = (
+                f'<div class="audio-viewer">'
+                f'<audio controls preload="metadata" src="{stream_url}"></audio>'
+                f'</div>'
+            )
+        else:
+            html = (
+                f'<div class="video-viewer">'
+                f'<video controls playsinline preload="metadata" src="{stream_url}"></video>'
+                f'</div>'
+            )
+        return {
+            "vault": vault_name,
+            "path": path,
+            "title": file_path.name,
+            "tags": [],
+            "frontmatter": {},
+            "html": html,
+            "raw_length": size,
+            "extension": ext,
+            "is_markdown": False,
+            "is_audio": media_kind == "audio",
+            "is_video": media_kind == "video",
+            "media_mime": mime,
+            "stream_url": stream_url,
+            "size_bytes": size,
+        }
+
+    try:
+        raw = file_path.read_text(encoding="utf-8", errors="replace")
+    except PermissionError as e:
+        logger.error(f"Permission denied reading file {path}: {e}")
+        raise HTTPException(status_code=403, detail=f"Permission denied: cannot read file {path}")
+    except UnicodeDecodeError:
+        # Binary / unsupported file — return structured info with download option
+        size = file_path.stat().st_size
+        return {
+            "vault": vault_name,
+            "path": path,
+            "title": file_path.name,
+            "tags": [],
+            "frontmatter": {},
+            "html": "",
+            "raw_length": size,
+            "extension": ext,
+            "is_markdown": False,
+            "unsupported": True,
+            "size_bytes": size,
+        }
+    except Exception as e:
+        logger.error(f"Unexpected error reading file {path}: {e}")
+        raise HTTPException(status_code=500, detail=f"Error reading file: {e!s}")
+
+    # === CSV: render as HTML table ===
+    if ext == ".csv":
+        import csv
+        import io as csv_io
+        reader = csv.reader(csv_io.StringIO(raw))
+        rows = list(reader)
+        if not rows:
+            html = "<p><em>Fichier CSV vide</em></p>"
+        else:
+            headers = rows[0]
+            data_rows = rows[1:]
+            html = '<div class="csv-table-wrapper"><table class="csv-table"><thead><tr>'
+            for h in headers:
+                html += f"<th>{h}</th>"
+            html += "</tr></thead><tbody>"
+            for row in data_rows:
+                html += "<tr>"
+                for cell in row:
+                    html += f"<td>{cell}</td>"
+                html += "</tr>"
+            html += "</tbody></table></div>"
+        return {
+            "vault": vault_name, "path": path,
+            "title": file_path.name, "tags": [], "frontmatter": {},
+            "html": html, "raw_length": len(raw), "extension": ext,
+            "is_markdown": False, "is_csv": True,
+        }
+
+    # === JSON: syntax-highlighted display ===
+    if ext == ".json":
+        import json as json_mod
+        try:
+            parsed = json_mod.loads(raw)
+            formatted = json_mod.dumps(parsed, indent=2, ensure_ascii=False)
+        except json_mod.JSONDecodeError:
+            formatted = raw
+        html = f"<pre class='json-viewer'><code>{html_mod.escape(formatted)}</code></pre>"
+        return {
+            "vault": vault_name, "path": path,
+            "title": file_path.name, "tags": [], "frontmatter": {},
+            "html": html, "raw_length": len(raw), "extension": ext,
+            "is_markdown": False, "is_json": True,
+        }
+
+    # === Excalidraw .excalidraw.md (Obsidian plugin format) ===
+    if path.lower().endswith(".excalidraw.md"):
+        import re as re_mod
+        raw_lower = file_path.read_text(encoding="utf-8", errors="replace")
+        # Check for excalidraw-plugin in frontmatter or body
+        if "excalidraw-plugin:" in raw_lower:
+            # Extract compressed JSON block
+            match = re_mod.search(r'```compressed-json\n(.*?)\n```', raw_lower, re_mod.DOTALL)
+            if match:
+                compressed = match.group(1).strip()
+                return {
+                    "vault": vault_name, "path": path,
+                    "title": file_path.name.replace(".excalidraw.md", ""),
+                    "tags": [], "frontmatter": {},
+                    "html": "", "raw_length": len(raw_lower),
+                    "extension": ".excalidraw.md",
+                    "is_markdown": False,
+                    "is_excalidraw": True,
+                    "excalidraw_data_compressed": compressed,
+                }
+        # Fallback: treat as regular markdown
+        raw = raw_lower
+    if ext == ".excalidraw":
+        import json as json_mod
+        try:
+            parsed = json_mod.loads(raw)
+        except json_mod.JSONDecodeError:
+            parsed = None
+        if parsed and parsed.get("type") == "excalidraw":
+            return {
+                "vault": vault_name,
+                "path": path,
+                "title": parsed.get("appState", {}).get("name") or file_path.name,
+                "tags": [],
+                "frontmatter": {},
+                "html": "",
+                "raw_length": len(raw),
+                "extension": ext,
+                "is_markdown": False,
+                "is_excalidraw": True,
+                "excalidraw_data": {
+                    "elements": parsed.get("elements", []),
+                    "appState": parsed.get("appState", {}),
+                    "files": parsed.get("files", {}),
+                },
+            }
+        else:
+            # Not a valid Excalidraw file — fall through to text viewer
+            pass
+
+    # === Plain text / other readable files ===
+    TEXT_EXTENSIONS = {".txt", ".log", ".yml", ".yaml", ".toml", ".ini", ".cfg",
+                       ".sh", ".bash", ".py", ".js", ".ts", ".html", ".css",
+                       ".xml", ".rst", ".tex", ".sql", ".conf", ".env"}
+    if ext in TEXT_EXTENSIONS or ext == ".md":
+        pass  # handled below or by markdown section
+
+    if ext == ".md":
+        post = parse_markdown_file(raw)
+
+        # Extract metadata using shared indexer logic
+        tags = _extract_tags(post)
+
+        title = post.metadata.get("title", file_path.stem.replace("-", " ").replace("_", " "))
+        html_content = _render_markdown(post.content, vault_name, file_path)
+
+        return {
+            "vault": vault_name,
+            "path": path,
+            "title": str(title),
+            "tags": tags,
+            "frontmatter": dict(post.metadata) if post.metadata else {},
+            "html": html_content,
+            "raw_length": len(raw),
+            "extension": ext,
+            "is_markdown": True,
+        }
+    else:
+        # Non-markdown: wrap in syntax-highlighted code block
+        lang = EXT_TO_LANG.get(ext, "")
+        if not lang:
+            # Fichiers sans extension usuels (Dockerfile, Makefile, etc.)
+            NAME_TO_LANG = {
+                "dockerfile": "dockerfile", "makefile": "makefile",
+                "cmakelists.txt": "cmake", "jenkinsfile": "groovy",
+                "vagrantfile": "ruby", "rakefile": "ruby", "gemfile": "ruby",
+                "procfile": "plaintext", "bashrc": "bash", "bash_profile": "bash",
+                "zshrc": "bash", "profile": "bash", "gitignore": "plaintext",
+            }
+            lang = NAME_TO_LANG.get(file_path.name.lower(), "plaintext")
+        escaped = html_mod.escape(raw)
+        html_content = f'<pre><code class="language-{lang}">{escaped}</code></pre>'
+
+        return {
+            "vault": vault_name,
+            "path": path,
+            "title": file_path.name,
+            "tags": [],
+            "frontmatter": {},
+            "html": html_content,
+            "raw_length": len(raw),
+            "extension": ext,
+            "is_markdown": False,
+        }
@@ -0,0 +1,532 @@
+"""File & directory mutation endpoints (ROADMAP #85, tranche 6b).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``PUT/DELETE/PATCH/POST /api/file/*``,
+``/api/directory/*``, ``/api/move/*``, ``/api/vault/*/batch-upload``),
+mêmes modèles de requête/réponse (déménagés dans :mod:`backend.schemas`),
+mêmes dépendances d'authentification et mêmes effets de bord (audit, index
+incrémental, SSE, webhooks, plugins, historique).
+
+La logique métier vit déjà dans :mod:`backend.services.mutations`.
+"""
+
+import logging
+from typing import Any
+
+from fastapi import APIRouter, Body, Depends, HTTPException, Query
+
+from backend.audit import log_file_delete, log_file_save
+from backend.auth.middleware import check_vault_access, require_auth
+from backend.history import (
+    remove_recent,
+    update_bookmarks_after_rename,
+    update_history_after_rename,
+)
+from backend.indexer import handle_file_move, remove_single_file, update_single_file
+from backend.schemas import (
+    BatchUploadRequest,
+    BatchUploadResponse,
+    DirectoryCreateRequest,
+    DirectoryCreateResponse,
+    DirectoryDeleteResponse,
+    DirectoryRenameRequest,
+    DirectoryRenameResponse,
+    FileCreateRequest,
+    FileCreateResponse,
+    FileDeleteResponse,
+    FileMoveRequest,
+    FileMoveResponse,
+    FileRenameRequest,
+    FileRenameResponse,
+    FileSaveResponse,
+)
+from backend.services.mutations import (
+    batch_upload_files as service_batch_upload_files,
+)
+from backend.services.mutations import (
+    create_directory as service_create_directory,
+)
+from backend.services.mutations import (
+    create_file as service_create_file,
+)
+from backend.services.mutations import (
+    delete_directory as service_delete_directory,
+)
+from backend.services.mutations import (
+    delete_file as service_delete_file,
+)
+from backend.services.mutations import (
+    edit_file as service_edit_file,
+)
+from backend.services.mutations import (
+    edit_xlsx_cells as service_edit_xlsx_cells,
+)
+from backend.services.mutations import (
+    move_path as service_move_path,
+)
+from backend.services.mutations import (
+    rename_directory as service_rename_directory,
+)
+from backend.services.mutations import (
+    rename_file as service_rename_file,
+)
+from backend.share import update_shares_after_rename
+from backend.sse import sse_manager
+from backend.webhooks import dispatch_webhooks
+
+logger = logging.getLogger("obsigate")
+
+router = APIRouter(tags=["files"])
+
+
+@router.put("/api/file/{vault_name}/save", response_model=FileSaveResponse)
+async def api_file_save(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to file"),
+    body: dict = Body(...),
+    backup: bool = Query(True, description="Create a backup before saving (default true, set false for auto-save)"),
+    current_user=Depends(require_auth),
+):
+    """Save (overwrite) a file's content.
+
+    Expects a JSON body with a ``content`` key containing the new text.
+    The path is validated against traversal attacks before writing.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative file path within the vault.
+        body: JSON body with ``content`` string.
+
+    Returns:
+        ``FileSaveResponse`` confirming the write.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    content = body.get("content", "")
+    result = service_edit_file(vault_name, path, content, backup=backup)
+
+    # Audit log
+    client_ip = current_user.get("_request_ip", "unknown")
+    log_file_save(current_user["username"], vault_name, path, len(content), client_ip)
+
+    return {"status": "ok", "vault": result["vault"], "path": result["path"], "size": result["size"]}
+
+
+@router.put("/api/file/{vault_name}/xlsx/save", response_model=FileSaveResponse)
+def api_file_xlsx_save(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to the .xlsx file"),
+    body: dict = Body(
+        ...,
+        description=(
+            '{"sheet": str, "cells": {"A1": value}, '
+            '"allow_formula": false, "force": false}'
+        ),
+    ),
+    current_user=Depends(require_auth),
+):
+    """Apply cell edits to an .xlsx workbook.
+
+    Expects a JSON body with ``sheet`` and ``cells`` (A1 references to new
+    scalar values, max 500 per request) plus two optional boolean flags:
+
+    * ``allow_formula`` — keep values starting with ``=``/``@`` as real
+      formulas. Off by default (#153 A4): such a value is stored as text so a
+      later Excel session cannot execute it (DDE).
+    * ``force`` — write a workbook carrying features openpyxl cannot re-serialize
+      (slicers, form controls, connections, custom XML, signature, cached formula
+      results). Without it the call fails **409** ``xlsx_lossy_content`` and the
+      client asks the user to confirm (#153 A1).
+
+    A backup is created before the workbook is rewritten, and the new archive
+    swaps in atomically. Declared as a sync endpoint on purpose: the openpyxl
+    round-trip and the per-file lock wait (#153 A3) then run in the threadpool
+    instead of blocking the event loop.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    sheet = body.get("sheet")
+    cells = body.get("cells")
+    if not isinstance(sheet, str) or not sheet:
+        raise HTTPException(status_code=400, detail="Feuille manquante")
+    if not isinstance(cells, dict) or not cells or len(cells) > 500:
+        raise HTTPException(status_code=400, detail="Cellules invalides (1 à 500 par requête)")
+    for ref, value in cells.items():
+        if not isinstance(ref, str) or not isinstance(value, (str, int, float, bool, type(None))):
+            raise HTTPException(status_code=400, detail=f"Cellule invalide: {ref!r}")
+    flags: dict[str, bool] = {}
+    for name in ("allow_formula", "force"):
+        raw = body.get(name, False)
+        if not isinstance(raw, bool):
+            raise HTTPException(status_code=400, detail=f"Flag invalide: {name}")
+        flags[name] = raw
+
+    result = service_edit_xlsx_cells(
+        vault_name, path, sheet, cells, **flags
+    )
+    log_file_save(
+        current_user["username"], vault_name, path,
+        sum(len(str(v)) for v in cells.values()),
+        current_user.get("_request_ip", "unknown"),
+    )
+    return {"status": "ok", "vault": result["vault"], "path": result["path"], "size": result["size"]}
+
+
+@router.delete("/api/file/{vault_name}", response_model=FileDeleteResponse)
+async def api_file_delete(vault_name: str, path: str = Query(..., description="Relative path to file"), current_user=Depends(require_auth)):
+    """Delete a file from the vault.
+
+    The path is validated against traversal attacks before deletion.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative file path within the vault.
+
+    Returns:
+        ``FileDeleteResponse`` confirming the deletion.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_delete_file(vault_name, path)
+
+    # Audit log
+    client_ip = current_user.get("_request_ip", "unknown")
+    log_file_delete(current_user["username"], vault_name, path, client_ip)
+
+    # Update index
+    await remove_single_file(vault_name, path)
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("file_deleted", {
+        "vault": vault_name,
+        "path": path,
+    })
+
+    from backend.plugins import emit_file_deleted
+    emit_file_deleted(vault_name, path)
+
+    # Remove from recent files
+    remove_recent(current_user["username"], vault_name, path)
+
+    # Dispatch webhooks
+    await dispatch_webhooks("file_deleted", {"vault": vault_name, "path": path})
+
+    return {"status": "ok", "vault": result["vault"], "path": result["path"]}
+
+
+@router.post("/api/directory/{vault_name}", response_model=DirectoryCreateResponse)
+async def api_directory_create(
+    vault_name: str,
+    body: DirectoryCreateRequest,
+    current_user=Depends(require_auth),
+):
+    """Create a new directory in a vault.
+
+    Args:
+        vault_name: Name of the vault.
+        body: Request body with directory path.
+
+    Returns:
+        DirectoryCreateResponse confirming creation.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_create_directory(vault_name, body.path)
+
+    # Update path_index with the new directory
+    from backend.indexer import _index_lock
+    from backend.indexer import path_index as _path_idx
+    with _index_lock:
+        if vault_name not in _path_idx:
+            _path_idx[vault_name] = []
+        existing = {p["path"] for p in _path_idx[vault_name]}
+        # Build all parent segments
+        parts = body.path.split("/")
+        for i in range(1, len(parts) + 1):
+            seg_path = "/".join(parts[:i])
+            if seg_path and seg_path not in existing:
+                existing.add(seg_path)
+                _path_idx[vault_name].append({
+                    "path": seg_path,
+                    "name": parts[i - 1],
+                    "type": "directory",
+                })
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("directory_created", {
+        "vault": vault_name,
+        "path": result["path"],
+    })
+    await dispatch_webhooks("directory_created", {"vault": vault_name, "path": result["path"]})
+
+    return {"success": True, "path": result["path"]}
+
+
+@router.patch("/api/directory/{vault_name}", response_model=DirectoryRenameResponse)
+async def api_directory_rename(
+    vault_name: str,
+    body: DirectoryRenameRequest,
+    current_user=Depends(require_auth),
+):
+    """Rename a directory in a vault.
+
+    Args:
+        vault_name: Name of the vault.
+        body: Request body with current path and new name.
+
+    Returns:
+        DirectoryRenameResponse with old and new paths.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_rename_directory(vault_name, body.path, body.new_name)
+    old_path_str = result["old_path"]
+    new_path_str = result["new_path"]
+
+    # Update index for all files in the directory
+    from backend.indexer import reload_single_vault
+    await reload_single_vault(vault_name)
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("directory_renamed", {
+        "vault": vault_name,
+        "old_path": old_path_str,
+        "new_path": new_path_str,
+    })
+    await dispatch_webhooks("directory_renamed", {"vault": vault_name, "old_path": old_path_str, "new_path": new_path_str})
+
+    return {"success": True, "old_path": old_path_str, "new_path": new_path_str}
+
+
+@router.delete("/api/directory/{vault_name}", response_model=DirectoryDeleteResponse)
+async def api_directory_delete(
+    vault_name: str,
+    path: str = Query(..., description="Relative path to directory"),
+    current_user=Depends(require_auth),
+):
+    """Delete a directory and all its contents from a vault.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative directory path within the vault.
+
+    Returns:
+        DirectoryDeleteResponse with count of deleted files.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_delete_directory(vault_name, path, recursive=True)
+    file_count = result["deleted_count"]
+
+    # Update index
+    from backend.indexer import reload_single_vault
+    await reload_single_vault(vault_name)
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("directory_deleted", {
+        "vault": vault_name,
+        "path": result["path"],
+        "deleted_count": file_count,
+    })
+    await dispatch_webhooks("directory_deleted", {"vault": vault_name, "path": result["path"]})
+
+    return {"success": True, "deleted_count": file_count}
+
+
+@router.post("/api/file/{vault_name}", response_model=FileCreateResponse)
+async def api_file_create(
+    vault_name: str,
+    body: FileCreateRequest,
+    current_user=Depends(require_auth),
+):
+    """Create a new file in a vault.
+
+    Args:
+        vault_name: Name of the vault.
+        body: Request body with file path and initial content.
+
+    Returns:
+        FileCreateResponse confirming creation.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_create_file(vault_name, body.path, body.content)
+
+    # Update index
+    await update_single_file(vault_name, result["path"])
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("file_created", {
+        "vault": vault_name,
+        "path": result["path"],
+    })
+    await dispatch_webhooks("file_created", {"vault": vault_name, "path": result["path"]})
+    from backend.plugins import emit_file_created
+    emit_file_created(vault_name, result["path"])
+
+    return {"success": True, "path": result["path"]}
+
+
+@router.post("/api/vault/{vault_name}/batch-upload", response_model=BatchUploadResponse)
+async def api_batch_upload(
+    vault_name: str,
+    body: BatchUploadRequest,
+    current_user=Depends(require_auth),
+):
+    """Upload multiple files and directories (recursively) into a vault.
+
+    Accepts base64 encoded or plain text files with relative directory paths.
+    Creates missing parent folders safely.
+
+    Args:
+        vault_name: Target vault name.
+        body: BatchUploadRequest with target_dir and files list.
+
+    Returns:
+        BatchUploadResponse with summary of uploaded files and errors.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    import base64
+
+    items: list[dict[str, Any]] = []
+    for f in body.files:
+        if f.is_dir:
+            items.append({"path": f.path, "is_dir": True})
+            continue
+
+        raw_bytes = b""
+        if f.content is not None:
+            # Check if content is base64 encoded data URI or raw base64
+            content_str = f.content
+            if content_str.startswith("data:") and ";base64," in content_str:
+                content_str = content_str.split(";base64,", 1)[1]
+            try:
+                raw_bytes = base64.b64decode(content_str)
+            except Exception:
+                # Fallback to utf-8 text encoding
+                raw_bytes = f.content.encode("utf-8")
+
+        items.append({"path": f.path, "content": raw_bytes, "is_dir": False})
+
+    result = service_batch_upload_files(
+        vault_name,
+        body.target_dir,
+        items,
+        overwrite=body.overwrite,
+    )
+
+    # Update index and SSE notifications for uploaded files
+    for path in result["uploaded"]:
+        try:
+            await update_single_file(vault_name, path)
+            await sse_manager.broadcast("file_created", {
+                "vault": vault_name,
+                "path": path,
+            })
+            await dispatch_webhooks("file_created", {"vault": vault_name, "path": path})
+        except Exception as e:
+            logger.warning(f"Failed to post-process upload of {path}: {e}")
+
+    # SSE notification for tree refresh
+    if result["uploaded"] or result["created_dirs"]:
+        await sse_manager.broadcast("tree_updated", {
+            "vault": vault_name,
+            "target_dir": result["target_dir"],
+        })
+
+    return result
+
+
+@router.patch("/api/file/{vault_name}", response_model=FileRenameResponse)
+async def api_file_rename(
+    vault_name: str,
+    body: FileRenameRequest,
+    current_user=Depends(require_auth),
+):
+    """Rename a file in a vault.
+
+    Args:
+        vault_name: Name of the vault.
+        body: Request body with current path and new name.
+
+    Returns:
+        FileRenameResponse with old and new paths.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_rename_file(vault_name, body.path, body.new_name)
+    old_path_str = result["old_path"]
+    new_path_str = result["new_path"]
+
+    # Update index
+    await handle_file_move(vault_name, old_path_str, new_path_str)
+
+    # Update bookmarks, history, and shares
+    update_bookmarks_after_rename(vault_name, old_path_str, new_path_str)
+    update_history_after_rename(vault_name, old_path_str, new_path_str)
+    update_shares_after_rename(vault_name, old_path_str, new_path_str)
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("file_renamed", {
+        "vault": vault_name,
+        "old_path": old_path_str,
+        "new_path": new_path_str,
+    })
+    await dispatch_webhooks("file_renamed", {"vault": vault_name, "old_path": old_path_str, "new_path": new_path_str})
+
+    return {"success": True, "old_path": old_path_str, "new_path": new_path_str}
+
+
+@router.post("/api/move/{vault_name}", response_model=FileMoveResponse)
+async def api_file_move(
+    vault_name: str,
+    body: FileMoveRequest,
+    current_user=Depends(require_auth),
+):
+    """Move a file or directory to a different parent directory within the same vault.
+
+    Supports both files and directories. The item keeps its original name;
+    only the parent directory changes.
+
+    Args:
+        vault_name: Name of the vault.
+        body: Request body with source_path and destination_dir.
+
+    Returns:
+        FileMoveResponse with old and new paths.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    result = service_move_path(vault_name, body.source_path, body.destination_dir)
+    old_path_str = result["old_path"]
+    new_path_str = result["new_path"]
+    item_type = result["item_type"]
+
+    # Update index
+    if item_type == "directory":
+        from backend.indexer import reload_single_vault
+        await reload_single_vault(vault_name)
+    else:
+        await handle_file_move(vault_name, old_path_str, new_path_str)
+
+    # Broadcast SSE event
+    await sse_manager.broadcast("item_moved", {
+        "vault": vault_name,
+        "old_path": old_path_str,
+        "new_path": new_path_str,
+        "item_type": item_type,
+    })
+    await dispatch_webhooks("item_moved", {"vault": vault_name, "old_path": old_path_str, "new_path": new_path_str, "item_type": item_type})
+
+    return {"success": True, "old_path": old_path_str, "new_path": new_path_str, "item_type": item_type}
@@ -0,0 +1,143 @@
+"""System health endpoints (ROADMAP #85, tranche 1).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/health``, ``/api/health/detailed``),
+même ``response_model`` (:class:`backend.schemas.HealthResponse`), même
+dépendance admin. Seule différence : la version est lue via
+:func:`backend.version.get_version` au lieu de ``app.version`` (valeur
+identique, figée au démarrage depuis le fichier ``VERSION``).
+
+Note : ``uptime_seconds`` reprend l'expression d'origine
+(``'_SERVER_START_TIME' in globals()``), qui vaut toujours 0 — le global
+n'est défini nulle part dans ``backend.main`` (voir ``backend.admin`` qui
+possède son propre compteur). Ce comportement est préservé tel quel ; le
+corriger fera l'objet d'une tranche ultérieure avec test dédié.
+"""
+
+from fastapi import APIRouter, Depends
+
+from backend.auth.middleware import require_admin
+from backend.indexer import index
+from backend.schemas import HealthResponse
+from backend.version import get_git_commit, get_git_describe, get_version
+
+router = APIRouter(tags=["System"])
+
+
+@router.get("/api/health", response_model=HealthResponse)
+async def api_health():
+    """Health check endpoint for Docker and monitoring.
+
+    Returns:
+        Application status, version, vault count and total file count.
+    """
+    total_files = sum(len(v["files"]) for v in index.values())
+    total_tokens = sum(len(v.get("files", [])) * 1000 for v in index.values())  # rough approx
+    import time
+
+    from backend.indexer import _last_full_index_ts
+    # `_SERVER_START_TIME` n'existe dans aucun module (comportement d'origine
+    # préservé : uptime toujours 0 — voir docstring du module).
+    uptime = int(time.time() - _SERVER_START_TIME) if '_SERVER_START_TIME' in globals() else 0  # noqa: F821
+    return {
+        "status": "ok",
+        "version": get_version(),
+        "vaults": len(index),
+        "total_files": total_files,
+        "total_tokens": total_tokens,
+        "last_full_index_ts": _last_full_index_ts,
+        "uptime_seconds": uptime,
+        "git_describe": get_git_describe(),
+        "git_commit": get_git_commit(),
+    }
+
+
+@router.get("/api/health/detailed", response_model=HealthResponse)
+async def api_health_detailed(current_user=Depends(require_admin)):
+    """Detailed health check — admin only.
+
+    Returns enriched metrics including memory, disk, SSE connections, and backup stats.
+    """
+
+    import psutil
+
+    from backend.admin import _count_active_sessions, _get_disk_stats
+    from backend.indexer import _last_full_index_ts, index
+
+    total_files = sum(len(v["files"]) for v in index.values())
+    total_tokens = sum(len(v.get("files", [])) * 1000 for v in index.values())
+    import time
+    uptime = int(time.time() - _SERVER_START_TIME) if '_SERVER_START_TIME' in globals() else 0  # noqa: F821 — voir ci-dessus
+
+    # Memory
+    vm = psutil.virtual_memory()
+    mem_used_mb = round(vm.used / (1024 ** 2), 1)
+    mem_total_mb = round(vm.total / (1024 ** 2), 1)
+    mem_pct = round(vm.percent, 1)
+
+    # CPU
+    cpu_pct = psutil.cpu_percent(interval=None)
+
+    # Disk
+    disk_used_gb, disk_total_gb = _get_disk_stats()
+    disk_free_gb = round(disk_total_gb - disk_used_gb, 2)
+    disk_pct = round((disk_used_gb / disk_total_gb * 100) if disk_total_gb > 0 else 0, 1)
+
+    # SSE connections (approximation)
+    active_sessions = _count_active_sessions()
+
+    # Backups
+    from backend.admin import _scan_backups
+    backup_rows = _scan_backups()
+    total_backups = len(backup_rows)
+    total_backup_size_mb = round(sum(r["size"] for r in backup_rows) / (1024 ** 2), 2)
+    oldest_backup_age_days = 0.0
+    if backup_rows:
+        now_ts = int(time.time())
+        oldest_ts = min(r["timestamp"] for r in backup_rows)
+        oldest_backup_age_days = round((now_ts - oldest_ts) / 86400, 2)
+
+    # Index details
+    index_detail = {}
+    for name, data in index.items():
+        index_detail[name] = {
+            "file_count": len(data["files"]),
+            "tag_count": len(data.get("tags", [])),
+            "token_count_approx": len(data.get("files", [])) * 1000,
+        }
+
+    return {
+        "status": "ok",
+        "version": get_version(),
+        "vaults": len(index),
+        "total_files": total_files,
+        "total_tokens": total_tokens,
+        "last_full_index_ts": _last_full_index_ts,
+        "uptime_seconds": uptime,
+        "git_describe": get_git_describe(),
+        "git_commit": get_git_commit(),
+        # Enriched fields
+        "memory": {
+            "used_mb": mem_used_mb,
+            "total_mb": mem_total_mb,
+            "percent": mem_pct,
+        },
+        "cpu": {
+            "percent": cpu_pct,
+        },
+        "disk": {
+            "used_gb": disk_used_gb,
+            "total_gb": disk_total_gb,
+            "free_gb": disk_free_gb,
+            "percent": disk_pct,
+        },
+        "connections": {
+            "active_sse": active_sessions,
+        },
+        "backups": {
+            "total_count": total_backups,
+            "total_size_mb": total_backup_size_mb,
+            "oldest_age_days": oldest_backup_age_days,
+        },
+        "index": index_detail,
+    }
@@ -0,0 +1,129 @@
+"""Shared helpers for the file routers (ROADMAP #85, tranche 6a).
+
+Petites fonctions pures extraites de :mod:`backend.main` sans changement
+de comportement. Regroupées ici car utilisées par plusieurs routers
+(``files_read`` aujourd'hui, ``files_media`` / mutations ensuite) :
+- :func:`content_disposition` — aussi utilisée par ``_stream_file_with_range``
+  (resté dans ``main`` jusqu'à la tranche media).
+- :func:`media_max_inline_bytes` — aussi utilisée par ``/api/media``.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import os
+import re
+from pathlib import Path
+
+from fastapi import HTTPException, Request
+from fastapi.responses import FileResponse, StreamingResponse
+
+
+def content_disposition(disposition: str, filename: str) -> str:
+    """Build a header-safe Content-Disposition value.
+
+    HTTP header values must be ASCII. Unicode filenames are sent per
+    RFC 5987 via ``filename*`` (percent-encoded UTF-8) with a pure-ASCII
+    ``filename`` fallback. This avoids a UnicodeDecodeError / HTTP 500 when
+    the filename contains accented characters (e.g. 'Bière blonde…pdf').
+    """
+    from urllib.parse import quote
+    ascii_name = "".join(c for c in filename if c.isascii() and (c.isalnum() or c in " _-.")).strip() or "file"
+    ext = Path(filename).suffix
+    if ext and not Path(ascii_name).suffix:
+        ascii_name = ascii_name + ext
+    return f"{disposition}; filename=\"{ascii_name}\"; filename*=UTF-8''{quote(filename)}"
+
+
+def media_max_inline_bytes() -> int:
+    """Maximum size (bytes) for inline audio/video playback (roadmap #109-A3).
+
+    Configurable via ``OBSIGATE_MEDIA_MAX_INLINE_MB`` (default 500 MB). Files
+    above the limit are not streamed in the viewer (the UI falls back to the
+    download button), which keeps a single uvicorn worker from being pinned by
+    multi-gigabyte media. Invalid or non-positive values fall back to default.
+    """
+    default_mb = 500
+    raw = os.environ.get("OBSIGATE_MEDIA_MAX_INLINE_MB", "").strip()
+    if not raw:
+        return default_mb * 1024 * 1024
+    try:
+        mb = float(raw)
+    except ValueError:
+        return default_mb * 1024 * 1024
+    if mb <= 0:
+        return default_mb * 1024 * 1024
+    return int(mb * 1024 * 1024)
+
+
+def stream_file_with_range(file_path: Path, request: Request, media_type: str):
+    """Return a file response honouring the HTTP ``Range`` header (roadmap #109).
+
+    Extrait de :mod:`backend.main` (``_stream_file_with_range``) sans
+    changement de comportement. Shared by ``pdf/stream`` and ``/api/media``:
+    a plain :class:`FileResponse` with ``Accept-Ranges: bytes`` when no range
+    is requested, or a :class:`StreamingResponse` (206 Partial Content,
+    64 KiB chunks) for a valid single range. An unsatisfiable range yields
+    ``416`` with a ``Content-Range: bytes */size`` header.
+
+    Reads are offloaded to threads so the event loop is never blocked
+    (ASYNC230), matching the previous inline implementation.
+    """
+    file_size = file_path.stat().st_size
+    range_header = request.headers.get("range")
+    disposition = content_disposition("inline", file_path.name)
+
+    if range_header:
+        # Parse "bytes=start-end" (single range only; multi-range is not used by viewers)
+        m = re.match(r"bytes=(\d*)-(\d*)", range_header)
+        if not m:
+            raise HTTPException(status_code=416,
+                                headers={"Content-Range": f"bytes */{file_size}"})
+        start_s, end_s = m.group(1), m.group(2)
+        if start_s == "" and end_s == "":
+            raise HTTPException(status_code=416,
+                                headers={"Content-Range": f"bytes */{file_size}"})
+        if start_s == "":
+            # suffix range: last N bytes
+            length = min(int(end_s), file_size)
+            start = file_size - length
+            end = file_size - 1
+        else:
+            start = int(start_s)
+            end = int(end_s) if end_s else file_size - 1
+            end = min(end, file_size - 1)
+        if start > end or start >= file_size:
+            raise HTTPException(status_code=416,
+                                headers={"Content-Range": f"bytes */{file_size}"})
+
+        chunk_size = end - start + 1
+
+        async def _partial():
+            f = await asyncio.to_thread(open, str(file_path), "rb")
+            try:
+                await asyncio.to_thread(f.seek, start)
+                remaining = chunk_size
+                while remaining > 0:
+                    data = await asyncio.to_thread(f.read, min(64 * 1024, remaining))
+                    if not data:
+                        break
+                    remaining -= len(data)
+                    yield data
+            finally:
+                await asyncio.to_thread(f.close)
+
+        return StreamingResponse(
+            _partial(),
+            status_code=206,
+            media_type=media_type,
+            headers={
+                "Content-Range": f"bytes {start}-{end}/{file_size}",
+                "Accept-Ranges": "bytes",
+                "Content-Length": str(chunk_size),
+                "Content-Disposition": disposition,
+            },
+        )
+
+    return FileResponse(str(file_path), media_type=media_type, headers={
+        "Accept-Ranges": "bytes",
+        "Content-Disposition": disposition})
@@ -0,0 +1,160 @@
+"""History endpoints — recent, bookmarks, saved searches (ROADMAP #85, tranche 8).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins, mêmes modèles (``BookmarkToggleRequest``
+déménagé dans :mod:`backend.schemas`), mêmes dépendances
+d'authentification.
+
+Adaptations strictement équivalentes :
+- ``_resolve_safe_path`` / ``_backup_file`` → :mod:`backend.services.paths`
+  et :mod:`backend.services.backups` (pass-through).
+- ``_load_config`` vient de :mod:`backend.routers.config`.
+"""
+
+import logging
+from pathlib import Path
+
+import frontmatter
+from fastapi import APIRouter, Body, Depends, HTTPException, Query
+
+from backend.auth.middleware import check_vault_access, require_auth
+from backend.history import get_bookmarks, toggle_bookmark
+from backend.indexer import find_file_in_index, get_vault_data, update_single_file
+from backend.routers.config import _load_config
+from backend.saved_searches import delete_saved, get_saved, save_search
+from backend.schemas import (
+    BookmarksResponse,
+    BookmarkToggleRequest,
+    BookmarkToggleResponse,
+    RecentResponse,
+    SavedSearch,
+    StatusResponse,
+)
+from backend.services.backups import create_backup
+from backend.services.paths import resolve_safe_path
+from backend.services.recent import humanize_mtime, list_recent
+
+logger = logging.getLogger("obsigate")
+
+router = APIRouter(tags=["Bookmarks"])
+
+
+@router.get("/api/recent", response_model=RecentResponse)
+async def api_recent(limit: int | None = Query(None), vault: str | None = Query(None), mode: str | None = Query("opened"), current_user=Depends(require_auth)):
+    config = _load_config()
+    actual_limit = limit if limit is not None else config.get("recent_files_limit", 20)
+
+    username = current_user.get("username")
+    user_vaults = current_user.get("_token_vaults") or current_user.get("vaults", [])
+
+    return list_recent(
+        username,
+        user_vaults,
+        vault=vault,
+        limit=actual_limit,
+        mode=mode or "opened",
+    )
+
+
+@router.get("/api/bookmarks", response_model=BookmarksResponse)
+async def api_bookmarks(vault: str | None = Query(None), current_user=Depends(require_auth)):
+    username = current_user.get("username")
+    user_vaults = current_user.get("_token_vaults") or current_user.get("vaults", [])
+
+    if not username:
+        return {"files": []}
+
+    history = get_bookmarks(username, vault_filter=vault)
+    files_resp = []
+    for item in history:
+        v_name = item["vault"]
+        if "*" not in user_vaults and v_name not in user_vaults:
+            continue
+
+        # Find in index to get metadata
+        f_idx = find_file_in_index(item["path"], v_name)
+        if f_idx:
+            files_resp.append({
+                "path": f_idx["path"],
+                "title": f_idx.get("title") or item["path"].split("/")[-1],
+                "vault": v_name,
+                "mtime": item["bookmarked_at"],
+                "mtime_human": humanize_mtime(item["bookmarked_at"]),
+                "size_bytes": f_idx.get("size", 0),
+                "tags": [f"#{t}" for t in f_idx.get("tags", [])][:5],
+                "bookmarked": True
+            })
+        else:
+            files_resp.append({
+                "path": item["path"],
+                "title": item.get("title") or item["path"].split("/")[-1],
+                "vault": v_name,
+                "mtime": item["bookmarked_at"],
+                "mtime_human": humanize_mtime(item["bookmarked_at"]),
+                "tags": [],
+                "bookmarked": True
+            })
+    return {
+        "files": files_resp,
+        "total": len(files_resp)
+    }
+
+
+@router.post("/api/bookmarks/toggle", response_model=BookmarkToggleResponse)
+async def api_toggle_bookmark(req: BookmarkToggleRequest, current_user=Depends(require_auth)):
+    username = current_user.get("username")
+    if not username:
+        raise HTTPException(status_code=401, detail="Not authenticated")
+
+    # Check vault access
+    if not check_vault_access(req.vault, current_user):
+        raise HTTPException(status_code=403, detail="Access denied to vault")
+
+    is_now_bookmarked = toggle_bookmark(username, req.vault, req.path, req.title or "")
+
+    # Update the file's YAML frontmatter: favoris: true/false
+    vault_data = get_vault_data(req.vault)
+    if vault_data:
+        file_path = resolve_safe_path(Path(vault_data["path"]), req.path)
+        if file_path.exists() and file_path.suffix == ".md":
+            try:
+                raw = file_path.read_text(encoding="utf-8", errors="replace")
+                post = frontmatter.loads(raw)
+                if is_now_bookmarked:
+                    post.metadata["favoris"] = True
+                elif "favoris" in post.metadata:
+                    del post.metadata["favoris"]
+                new_raw = frontmatter.dumps(post)
+                create_backup(file_path, req.vault, req.path)
+                file_path.write_text(new_raw, encoding="utf-8")
+                await update_single_file(req.vault, str(file_path))
+            except Exception as e:
+                logger.warning(f"Failed to update favoris metadata on {req.vault}/{req.path}: {e}")
+
+    return {"bookmarked": is_now_bookmarked}
+
+
+@router.get("/api/saved-searches", response_model=list[SavedSearch])
+async def api_saved_searches(current_user=Depends(require_auth)):
+    username = current_user.get("username")
+    if not username:
+        raise HTTPException(401)
+    return get_saved(username)
+
+
+@router.post("/api/saved-searches", response_model=SavedSearch)
+async def api_save_search(body: dict = Body(...), current_user=Depends(require_auth)):
+    username = current_user.get("username")
+    if not username:
+        raise HTTPException(401)
+    return save_search(username, body)
+
+
+@router.delete("/api/saved-searches/{search_id}", response_model=StatusResponse)
+async def api_delete_saved_search(search_id: str, current_user=Depends(require_auth)):
+    username = current_user.get("username")
+    if not username:
+        raise HTTPException(401)
+    if not delete_saved(username, search_id):
+        raise HTTPException(404, "Not found")
+    return {"status": "deleted"}
@@ -0,0 +1,105 @@
+"""Real-time endpoints — SSE stream & collaboration WebSocket (ROADMAP #85, tranche 9).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/events``,
+``/ws/collab/{vault}/{path}``), même authentification (Depend pour le SSE,
+manuelle pour le WebSocket — les ``Depends`` FastAPI ne s'exécutent pas sur
+les routes WebSocket).
+
+Pas de tags déclarés : assignation par chemin via
+``openapi_docs.tag_for_path`` comme avant (``/api/events`` → System).
+"""
+
+import asyncio
+import json as _json
+
+from fastapi import APIRouter, Depends, WebSocket
+from fastapi.responses import StreamingResponse
+
+from backend.auth.middleware import check_vault_access, require_auth
+from backend.collab import authenticate_websocket, collab_manager
+from backend.services.paths import resolve_safe_path
+from backend.services.vaults import get_vault_root
+from backend.sse import sse_manager
+
+router = APIRouter()
+
+
+@router.get(
+    "/api/events",
+    response_class=StreamingResponse,
+    responses={200: {"content": {"text/event-stream": {}}, "description": "Server-Sent Events stream"}},
+)
+async def api_events(current_user=Depends(require_auth)):
+    """SSE stream for real-time index update notifications.
+
+    Sends keepalive comments every 30s. Events:
+    - ``index_updated``: partial index change (file create/modify/delete/move)
+    - ``index_reloaded``: full re-index completed
+    - ``vault_added``: new vault added dynamically
+    - ``vault_removed``: vault removed dynamically
+    """
+    queue = await sse_manager.connect()
+
+    async def event_generator():
+        try:
+            # Send initial connection event
+            yield f"event: connected\ndata: {_json.dumps({'sse_clients': sse_manager.client_count})}\n\n"
+            while True:
+                try:
+                    msg = await asyncio.wait_for(queue.get(), timeout=30.0)
+                    yield f"event: {msg['event']}\ndata: {msg['data']}\n\n"
+                except asyncio.TimeoutError:
+                    # Keepalive comment
+                    yield ": keepalive\n\n"
+                except asyncio.CancelledError:
+                    break
+        finally:
+            sse_manager.disconnect(queue)
+
+    return StreamingResponse(
+        event_generator(),
+        media_type="text/event-stream",
+        headers={
+            "Cache-Control": "no-cache",
+            "Connection": "keep-alive",
+            "X-Accel-Buffering": "no",
+        },
+    )
+
+
+@router.websocket("/ws/collab/{vault_name}/{path:path}")
+async def collab_websocket(websocket: WebSocket, vault_name: str, path: str):
+    """Real-time collaborative editing over WebSocket (ROADMAP #62).
+
+    One *room* is created per ``vault::path``; all clients editing the same
+    file share Yjs/CRDT updates, awareness (cursors/selection) and a debounced
+    server-side persistence of the markdown content.
+
+    Authentication is performed manually (FastAPI ``Depends`` do not run for
+    WebSocket routes) and vault access is enforced per connection.
+    """
+    from backend.services.errors import ServiceError
+
+    user = authenticate_websocket(websocket)
+    if user is None:
+        await websocket.close(code=4401)
+        return
+
+    if not check_vault_access(vault_name, user):
+        await websocket.close(code=4403)
+        return
+
+    try:
+        vault_root = get_vault_root(vault_name)
+        file_path = resolve_safe_path(vault_root, path)
+    except ServiceError:
+        await websocket.close(code=4404)
+        return
+
+    if not file_path.exists() or not file_path.is_file():
+        await websocket.close(code=4404)
+        return
+
+    await websocket.accept()
+    await collab_manager.connect(websocket, vault_name, path, file_path, user)
@@ -0,0 +1,353 @@
+"""Search, suggest, graph & index-reload endpoints (ROADMAP #85, tranche 5).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins, mêmes modèles de réponse (déménagés dans
+:mod:`backend.schemas`), mêmes dépendances d'authentification. La logique
+métier vit déjà dans :mod:`backend.services.search`,
+:mod:`backend.search`, :mod:`backend.services.graph` et
+:mod:`backend.services.mutations`.
+
+Adaptations strictement équivalentes :
+- Le pool ``_search_executor`` de ``main`` vit désormais dans
+  :mod:`backend.search_executor` (même dimensionnement, même cycle de vie
+  géré par le lifespan de ``main``) : accès via
+  :func:`get_search_executor`.
+"""
+
+import asyncio
+import logging
+from functools import partial
+from pathlib import Path
+
+from fastapi import APIRouter, Body, Depends, HTTPException, Query
+
+from backend.audit import log_file_save
+from backend.auth.middleware import check_vault_access, require_admin, require_auth
+from backend.indexer import get_vault_data, reload_index, update_single_file
+from backend.schemas import (
+    AdvancedSearchResponse,
+    GraphResponse,
+    ReloadResponse,
+    ReplaceResponse,
+    SearchResponse,
+    SuggestResponse,
+    TagsResponse,
+    TagSuggestResponse,
+    TreeSearchResponse,
+    VaultPathsResponse,
+    VaultStatsResponse,
+)
+from backend.search import suggest_tags, suggest_titles
+from backend.search_executor import get_search_executor
+from backend.services.graph import get_graph as service_get_graph
+from backend.services.mutations import (
+    replace_in_files as service_replace_in_files,
+)
+from backend.services.search import advanced_search_vaults, list_paths, search_paths, search_vaults
+from backend.services.search import list_tags as service_list_tags
+from backend.sse import sse_manager
+
+logger = logging.getLogger("obsigate")
+
+router = APIRouter(tags=["search"])
+
+
+@router.get("/api/search", response_model=SearchResponse)
+async def api_search(
+    q: str = Query("", description="Search query"),
+    vault: str = Query("all", description="Vault filter"),
+    tag: str | None = Query(None, description="Tag filter"),
+    limit: int = Query(50, ge=1, le=200, description="Results per page"),
+    offset: int = Query(0, ge=0, description="Pagination offset"),
+    current_user=Depends(require_auth),
+):
+    """Full-text search across vaults with relevance scoring.
+
+    Supports combining free-text queries with tag filters.
+    Results are ranked by a multi-factor scoring algorithm.
+    Pagination via ``limit`` and ``offset`` (defaults preserve backward compat).
+
+    Args:
+        q: Free-text search string.
+        vault: Vault name or ``"all"`` to search everywhere.
+        tag: Comma-separated tag names to require.
+        limit: Max results per page (1–200).
+        offset: Pagination offset.
+
+    Returns:
+        ``SearchResponse`` with ranked results and snippets.
+    """
+    loop = asyncio.get_event_loop()
+    # Fetch the full result set (capped at DEFAULT_SEARCH_LIMIT internally) and
+    # paginate in the shared service so routes and tools share the same logic.
+    return await loop.run_in_executor(
+        get_search_executor(),
+        partial(search_vaults, q, vault, tag, limit, offset),
+    )
+
+
+@router.get("/api/tags", response_model=TagsResponse)
+async def api_tags(vault: str | None = Query(None, description="Vault filter"), current_user=Depends(require_auth)):
+    """Return all unique tags with occurrence counts.
+
+    Args:
+        vault: Optional vault name to restrict tag aggregation.
+
+    Returns:
+        ``TagsResponse`` with tags sorted by descending count.
+    """
+    return {"vault_filter": vault, "tags": service_list_tags(vault)}
+
+
+@router.get("/api/tree-search", response_model=TreeSearchResponse)
+async def api_tree_search(
+    q: str = Query("", description="Search query"),
+    vault: str = Query("all", description="Vault filter"),
+    current_user=Depends(require_auth),
+):
+    """Search for files and directories in the tree structure using pre-built index.
+
+    Uses the in-memory path index for instant filtering without filesystem access.
+
+    Args:
+        q: Search string to match against file/directory paths.
+        vault: Vault name or "all" to search everywhere.
+
+    Returns:
+        ``TreeSearchResponse`` with matching paths.
+    """
+    return search_paths(q, vault)
+
+
+@router.get("/api/vault/{vault_name}/paths", response_model=VaultPathsResponse)
+async def api_vault_paths(
+    vault_name: str,
+    limit: int = Query(5000, ge=1, le=20000, description="Maximum number of indexed paths to return"),
+    current_user=Depends(require_auth),
+):
+    """Return a flat list of every indexed file and directory in a vault.
+
+    Used by the AI assistant ``@`` mention menu to filter paths instantly on
+    the client (one request instead of one per keystroke).
+
+    Args:
+        vault_name: Name of the vault.
+        limit: Maximum number of entries returned.
+
+    Returns:
+        ``VaultPathsResponse`` with the vault's indexed paths.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+    return list_paths(vault_name, limit=limit)
+
+
+@router.get("/api/search/advanced", response_model=AdvancedSearchResponse)
+async def api_advanced_search(
+    q: str = Query("", description="Advanced search query (supports tag:, vault:, title:, path:, ext: operators)"),
+    vault: str = Query("all", description="Vault filter"),
+    tag: str | None = Query(None, description="Comma-separated tag filter"),
+    limit: int = Query(50, ge=1, le=200, description="Results per page"),
+    offset: int = Query(0, ge=0, description="Pagination offset"),
+    sort: str = Query("relevance", description="Sort by 'relevance' or 'modified'"),
+    case_sensitive: bool = Query(False, description="Match case"),
+    whole_word: bool = Query(False, description="Match whole words only"),
+    regex: bool = Query(False, description="Treat query as regex"),
+    include_paths: str | None = Query(None, description="Comma-separated glob patterns to include"),
+    exclude_paths: str | None = Query(None, description="Comma-separated glob patterns to exclude"),
+    created: str | None = Query(None, description="Created date filter (>date, <date, date..date)"),
+    modified: str | None = Query(None, description="Modified date filter (>date, <date, date..date, <Nd)"),
+    size: str | None = Query(None, description="Size filter (>size, <size, size..size, e.g. >1MB, <10KB)"),
+    semantic: bool = Query(False, description="Fuse TF-IDF with semantic embeddings (RRF)"),
+    current_user=Depends(require_auth),
+):
+    """Advanced full-text search with TF-IDF scoring, facets, and pagination.
+
+    Supports advanced query operators:
+    - ``tag:<name>`` or ``#<name>`` — filter by tag
+    - ``vault:<name>`` — filter by vault
+    - ``title:<text>`` — filter by title substring
+    - ``path:<text>`` — filter by path substring
+    - ``ext:<type>`` — filter by file extension
+    - ``created:>2024-01-01`` — filter by creation date
+    - ``modified:<7d`` or ``modified:2024-01-01..2024-06-01`` — filter by modification date
+    - ``size:>1MB`` or ``size:100KB..1MB`` — filter by file size
+    - Remaining text is scored using TF-IDF with accent normalization.
+    - Toggles: case_sensitive, whole_word, regex
+    - Path filters: include_paths, exclude_paths (glob patterns)
+    - ``semantic=true`` — fuse the TF-IDF ranking with the semantic (embedding)
+      ranking via Reciprocal Rank Fusion and expose ``semantic_score`` per result.
+
+    Results include ``<mark>``-highlighted snippets and faceted tag/vault counts.
+    """
+    loop = asyncio.get_event_loop()
+    search_fn = partial(advanced_search_vaults, q, vault=vault, tag=tag,
+                limit=limit, offset=offset, sort=sort,
+                case_sensitive=case_sensitive, whole_word=whole_word, regex=regex,
+                include_paths=include_paths, exclude_paths=exclude_paths,
+                created=created, modified=modified, size=size, semantic=semantic)
+    try:
+        return await loop.run_in_executor(get_search_executor(), search_fn)
+    except ValueError as e:
+        raise HTTPException(400, str(e)) from e
+
+
+@router.post("/api/search/replace", response_model=ReplaceResponse)
+async def api_search_replace(
+    body: dict = Body(...),
+    current_user=Depends(require_auth),
+):
+    """Find and replace across vault files."""
+    query = body.get("query", "")
+    replacement = body.get("replacement", "")
+    vault_filter = body.get("vault", "all")
+    case_sensitive = body.get("case_sensitive", False)
+    whole_word = body.get("whole_word", False)
+    regex_mode = body.get("regex", False)
+    include_paths = body.get("include_paths")
+    exclude_paths = body.get("exclude_paths")
+    replace_all = body.get("replace_all", False)
+    dry_run = body.get("dry_run", not replace_all)
+
+    if not query:
+        raise HTTPException(400, "Query is required")
+
+    result = service_replace_in_files(
+        query,
+        replacement,
+        vault=vault_filter,
+        case_sensitive=case_sensitive,
+        whole_word=whole_word,
+        regex=regex_mode,
+        include_paths=include_paths,
+        exclude_paths=exclude_paths,
+        replace_all=replace_all,
+        dry_run=dry_run,
+        is_vault_allowed=lambda v: check_vault_access(v, current_user),
+    )
+
+    if dry_run:
+        return result
+
+    # Side effects for applied replacements (audit + incremental index).
+    for match in result.get("replaced", []):
+        log_file_save(current_user["username"], match["vault"], match["path"], match.get("size", 0))
+        vault_data = get_vault_data(match["vault"])
+        if vault_data:
+            abs_path = str(Path(vault_data["path"]) / match["path"])
+            await update_single_file(match["vault"], abs_path)
+
+    return result
+
+
+@router.get("/api/suggest", response_model=SuggestResponse)
+async def api_suggest(
+    q: str = Query("", description="Prefix to search for in file titles"),
+    vault: str = Query("all", description="Vault filter"),
+    limit: int = Query(10, ge=1, le=50, description="Max suggestions"),
+    current_user=Depends(require_auth),
+):
+    """Suggest file titles matching a prefix (accent-insensitive).
+
+    Used for autocomplete in the search input.
+
+    Args:
+        q: User-typed prefix (minimum 2 characters).
+        vault: Vault name or ``"all"``.
+        limit: Max number of suggestions.
+
+    Returns:
+        ``SuggestResponse`` with matching file title suggestions.
+    """
+    suggestions = suggest_titles(q, vault_filter=vault, limit=limit)
+    return {"query": q, "suggestions": suggestions}
+
+
+@router.get("/api/tags/suggest", response_model=TagSuggestResponse)
+async def api_tags_suggest(
+    q: str = Query("", description="Prefix to search for in tags"),
+    vault: str = Query("all", description="Vault filter"),
+    limit: int = Query(10, ge=1, le=50, description="Max suggestions"),
+    current_user=Depends(require_auth),
+):
+    """Suggest tags matching a prefix (accent-insensitive).
+
+    Used for autocomplete when typing ``tag:`` or ``#`` in the search input.
+
+    Args:
+        q: User-typed prefix (with or without ``#``, minimum 2 characters).
+        vault: Vault name or ``"all"``.
+        limit: Max number of suggestions.
+
+    Returns:
+        ``TagSuggestResponse`` with matching tag suggestions and counts.
+    """
+    suggestions = suggest_tags(q, vault_filter=vault, limit=limit)
+    return {"query": q, "suggestions": suggestions}
+
+
+@router.get("/api/index/reload", response_model=ReloadResponse)
+async def api_reload(current_user=Depends(require_admin)):
+    """Force a full re-index of all configured vaults.
+
+    Returns:
+        ``ReloadResponse`` with per-vault file and tag counts.
+    """
+    stats = await reload_index()
+    await sse_manager.broadcast("index_reloaded", {
+        "vaults": list(stats.keys()),
+        "stats": stats,
+    })
+    return {"status": "ok", "vaults": stats}
+
+
+@router.get("/api/graph/{vault_name}", response_model=GraphResponse)
+async def api_graph(
+    vault_name: str,
+    path: str = Query("", description="Relative path to focus on"),
+    depth: int = Query(1, ge=0, le=3, description="How many levels deep to expand"),
+    scope: str = Query("directory", description="'directory' (default) or 'full' for entire vault"),
+    tag: str = Query("", description="Filter: only show files with this tag"),
+    current_user=Depends(require_auth),
+):
+    """Return graph data (nodes and edges) for a vault or directory.
+
+    Nodes represent files and directories. Edges represent parent-child
+    relationships and wikilinks between markdown files.
+
+    Args:
+        vault_name: Name of the vault.
+        path: Relative directory path to focus on (empty = root).
+        depth: Expansion depth (0 = only direct children, 1-3 = deeper).
+        scope: 'directory' for subtree, 'full' for entire vault.
+        tag: Optional tag filter (only files with this tag appear).
+
+    Returns:
+        ``GraphResponse`` with nodes and edges.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(status_code=403, detail=f"Accès refusé à la vault '{vault_name}'")
+
+    return service_get_graph(vault_name, path=path, depth=depth, scope=scope, tag=tag)
+
+
+@router.get("/api/index/reload/{vault_name}", response_model=VaultStatsResponse)
+async def api_reload_vault(vault_name: str, current_user=Depends(require_admin)):
+    """Force a re-index of a single vault.
+
+    Args:
+        vault_name: Name of the vault to reindex.
+
+    Returns:
+        Dict with vault statistics.
+    """
+    try:
+        from backend.indexer import reload_single_vault
+        stats = await reload_single_vault(vault_name)
+        await sse_manager.broadcast("vault_reloaded", {
+            "vault": vault_name,
+            "stats": stats,
+        })
+        return {"status": "ok", "vault": vault_name, "stats": stats}
+    except ValueError as e:
+        raise HTTPException(status_code=404, detail=str(e))
@@ -0,0 +1,306 @@
+"""Public share endpoints (ROADMAP #85, tranche 3).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/share/*``, ``/api/shares``,
+``/s/{token}*``), mêmes modèles de réponse, mêmes dépendances
+d'authentification (les pages ``/s/*`` restent publiques). La logique
+métier vit déjà dans :mod:`backend.share`.
+
+Adaptations strictement équivalentes (pas de changement de comportement) :
+- ``_resolve_safe_path`` / ``_backup_file`` de ``main`` n'étaient que des
+  wrappers directs : appelés ici via :mod:`backend.services.paths` et
+  :mod:`backend.services.backups` (mêmes signatures, mêmes exceptions
+  ``ServiceError`` toujours mappées par le handler global de ``main``).
+- ``_render_markdown`` vient de :mod:`backend.render` (#85 T9, sans cycle
+  d'import).
+"""
+
+import html as html_mod
+import json as _json
+import logging
+from pathlib import Path
+
+import frontmatter
+from fastapi import APIRouter, Body, Depends, HTTPException, Query, Request
+from fastapi.responses import FileResponse, HTMLResponse, Response
+
+from backend.auth.middleware import check_vault_access, require_auth
+from backend.indexer import get_vault_data, parse_markdown_file, update_single_file
+from backend.render import _render_markdown
+from backend.schemas import ShareModel, StatusResponse
+from backend.secret_redactor import redact_file_content
+from backend.services.backups import create_backup
+from backend.services.paths import resolve_safe_path
+from backend.share import (
+    create_share,
+    get_share_by_token,
+    list_shares,
+    record_access,
+    revoke_share,
+)
+
+logger = logging.getLogger("obsigate")
+
+# Lazy import: WeasyPrint PDF export (requires GTK, may not be available everywhere)
+try:
+    from backend.pdf_export import build_pdf_html, generate_pdf
+except Exception:  # pragma: no cover - WeasyPrint/GTK missing
+    generate_pdf = None  # type: ignore[assignment]
+    build_pdf_html = None  # type: ignore[assignment]
+
+    logging.getLogger("obsigate").warning("PDF export unavailable (WeasyPrint/GTK not found)")
+
+router = APIRouter(tags=["sharing"])
+
+
+@router.post("/api/share/{vault_name}", response_model=ShareModel)
+async def api_share_create(
+    vault_name: str,
+    body: dict = Body(...),
+    current_user=Depends(require_auth),
+):
+    """Create a public share link for a document.
+
+    Also sets ``publish: true`` in the file's YAML frontmatter so the
+    frontend can visually indicate the file is publicly shared.
+    """
+    if not check_vault_access(vault_name, current_user):
+        raise HTTPException(403, f"Accès refusé à la vault '{vault_name}'")
+    path = body.get("path", "")
+    expires = body.get("expires_in_hours")
+    share = create_share(vault_name, path, current_user["username"], expires)
+    share["url"] = f"/s/{share['token']}"
+
+    # Set publish: true in the file's frontmatter
+    vault_data = get_vault_data(vault_name)
+    if vault_data:
+        file_path = resolve_safe_path(Path(vault_data["path"]), path)
+        if file_path.exists() and file_path.suffix == ".md":
+            try:
+                raw = file_path.read_text(encoding="utf-8", errors="replace")
+                post = frontmatter.loads(raw)
+                if not post.metadata.get("publish"):
+                    post.metadata["publish"] = True
+                    new_raw = frontmatter.dumps(post)
+                    create_backup(file_path, vault_name, path)
+                    file_path.write_text(new_raw, encoding="utf-8")
+                    await update_single_file(vault_name, str(file_path))
+                    logger.info(f"Set publish:true on {vault_name}/{path}")
+            except Exception as e:
+                logger.warning(f"Failed to set publish metadata on {vault_name}/{path}: {e}")
+
+    return share
+
+
+@router.get("/api/shares", response_model=list[ShareModel])
+async def api_shares_list(vault: str | None = Query(None), current_user=Depends(require_auth)):
+    """List all shares (optionally filtered by vault)."""
+    shares = list_shares(vault)
+    for s in shares:
+        s["url"] = f"/s/{s['token']}"
+    return shares
+
+
+@router.delete("/api/share/{share_id}", response_model=StatusResponse)
+async def api_share_revoke(share_id: str, current_user=Depends(require_auth)):
+    if not revoke_share(share_id):
+        raise HTTPException(404, "Share not found")
+    return {"status": "revoked"}
+
+
+@router.get(
+    "/s/{token}/pdf",
+    response_class=Response,
+    responses={200: {"content": {"application/pdf": {}}, "description": "Shared document as PDF"}},
+)
+async def public_share_pdf_download(token: str):
+    """Download shared document as real PDF via WeasyPrint."""
+    if generate_pdf is None:
+        raise HTTPException(501, "PDF export unavailable (WeasyPrint/GTK not available)")
+    share = get_share_by_token(token)
+    if not share:
+        raise HTTPException(404, "Share not found or expired")
+    vault_data = get_vault_data(share["vault"])
+    if not vault_data:
+        raise HTTPException(404, "Vault not found")
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, share["path"])
+    if not file_path.exists():
+        raise HTTPException(404, "File not found")
+    try:
+        raw = file_path.read_text(encoding="utf-8", errors="replace")
+    except Exception:
+        raise HTTPException(500, "Cannot read file")
+    record_access(token)
+    raw = redact_file_content(raw, str(file_path))
+    post = parse_markdown_file(raw)
+    ext = file_path.suffix.lower()
+    if ext == ".md":
+        html = _render_markdown(post.content, share["vault"], file_path)
+    else:
+        html = f'<pre style="font-family:monospace;font-size:12px;line-height:1.6;white-space:pre-wrap">{html_mod.escape(raw)}</pre>'
+    title = post.metadata.get("title", file_path.stem)
+    pdf_html = build_pdf_html(html, str(title))
+    pdf_bytes = generate_pdf(pdf_html, str(title))
+    safe_name = "".join(c for c in str(title) if c.isascii() and (c.isalnum() or c in " _-.")).strip() or "document"
+    return Response(content=pdf_bytes, media_type="application/pdf", headers={"Content-Disposition": f'attachment; filename="{safe_name}.pdf"'})
+
+
+@router.get("/s/{token}/raw", response_class=FileResponse)
+async def public_share_raw(token: str):
+    """Download the raw (original) shared document."""
+    share = get_share_by_token(token)
+    if not share:
+        raise HTTPException(404, "Share not found or expired")
+    vault_data = get_vault_data(share["vault"])
+    if not vault_data:
+        raise HTTPException(404, "Vault not found")
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, share["path"])
+    if not file_path.exists():
+        raise HTTPException(404, "File not found")
+    record_access(token)
+    return FileResponse(path=str(file_path), filename=file_path.name, media_type="application/octet-stream")
+
+
+@router.get("/s/{token}", response_class=HTMLResponse)
+async def public_share_view(request: Request, token: str):
+    """Public share view — no authentication required."""
+    from backend.csp import inject_csp_nonce
+
+    share = get_share_by_token(token)
+    if not share:
+        raise HTTPException(404, "Share not found or expired")
+    vault_data = get_vault_data(share["vault"])
+    if not vault_data:
+        raise HTTPException(404, "Vault not found")
+    vault_root = Path(vault_data["path"])
+    file_path = resolve_safe_path(vault_root, share["path"])
+    if not file_path.exists():
+        raise HTTPException(404, "File not found")
+    try:
+        raw = file_path.read_text(encoding="utf-8", errors="replace")
+    except Exception:
+        raise HTTPException(500, "Cannot read file")
+    record_access(token)
+    raw = redact_file_content(raw, str(file_path))
+    post = parse_markdown_file(raw)
+    ext = file_path.suffix.lower()
+
+    if ext == ".md":
+        html = _render_markdown(post.content, share["vault"], file_path)
+    else:
+        escaped = html_mod.escape(raw)
+        html = f'<pre style="background:var(--bg-card);border:1px solid var(--border);border-radius:8px;padding:16px;overflow-x:auto;font-size:0.85rem;line-height:1.6"><code>{escaped}</code></pre>'
+
+    title = post.metadata.get("title", file_path.stem)
+
+    # Escape everything user-controlled before embedding in HTML/JS (BUG-022).
+    title_esc = html_mod.escape(str(title))
+    # Neutralise ``</script>`` in the JS string literal too.
+    title_download_js = (
+        _json.dumps(f"{title}.md")
+        .replace("<", "\\u003c")
+        .replace(">", "\\u003e")
+        .replace("&", "\\u0026")
+    )
+
+    # JSON-escape raw content for embedding in HTML, and neutralise ``</script>``.
+    raw_json = (
+        _json.dumps(raw)
+        .replace("<", "\\u003c")
+        .replace(">", "\\u003e")
+        .replace("&", "\\u0026")
+    )
+    fm_html = ""
+    if post.metadata:
+        fm_items = []
+        skip_keys = {"title", "titre"}
+        for k, v in post.metadata.items():
+            if k in skip_keys:
+                continue
+            if isinstance(v, list):
+                v = ", ".join(str(x) for x in v)
+            elif isinstance(v, bool):
+                v = "✓" if v else "✗"
+            elif v is None:
+                v = "—"
+            fm_items.append(
+                f'<div class="fm-row"><span class="fm-key">{html_mod.escape(str(k))}</span>'
+                f'<span class="fm-val">{html_mod.escape(str(v))}</span></div>'
+            )
+        if fm_items:
+            fm_html = f'<div class="fm-section"><div class="fm-header">Frontmatter</div><div class="fm-body">{"".join(fm_items)}</div></div>'
+
+    return HTMLResponse(
+        inject_csp_nonce(
+            f"""<!DOCTYPE html><html lang="fr" data-theme="dark"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">
+<title>{title_esc} — ObsiGate Share</title>
+<style>
+:root {{ --bg:#1a1a2e; --bg-card:#16213e; --text:#e0e0e0; --text-muted:#888; --accent:#6366f1; --border:#2a2a4a; --banner-bg:var(--accent); --banner-text:#fff; }}
+[data-theme="light"] {{ --bg:#f8f9fa; --bg-card:#fff; --text:#1a1a2e; --text-muted:#666; --accent:#4f46e5; --border:#ddd; --banner-bg:#eef2ff; --banner-text:#4338ca; }}
+*{{box-sizing:border-box;margin:0;padding:0}}
+body{{font-family:system-ui,-apple-system,sans-serif;background:var(--bg);color:var(--text);line-height:1.7;min-height:100vh}}
+.toolbar{{position:sticky;top:0;z-index:10;background:var(--bg-card);border-bottom:1px solid var(--border);padding:8px 16px;display:flex;align-items:center;gap:8px;flex-wrap:wrap}}
+.toolbar-title{{font-weight:600;font-size:0.9rem;margin-right:auto;overflow:hidden;text-overflow:ellipsis;white-space:nowrap}}
+.toolbar-btn{{padding:6px 12px;border:1px solid var(--border);border-radius:6px;background:var(--bg);color:var(--text);cursor:pointer;font-size:0.8rem;display:flex;align-items:center;gap:5px;transition:all .15s}}
+.toolbar-btn:hover{{background:var(--accent);color:#fff;border-color:var(--accent)}}
+.toolbar-btn svg{{width:15px;height:15px;flex-shrink:0}}
+.toolbar-btn:hover svg{{stroke:#fff}}
+.share-banner{{background:var(--banner-bg);color:var(--banner-text);padding:6px 16px;font-size:0.8rem;text-align:center;display:flex;align-items:center;justify-content:center;gap:6px}}
+.share-banner svg{{width:14px;height:14px;flex-shrink:0}}
+.content{{max-width:820px;margin:0 auto;padding:24px 20px 60px}}
+.content h1{{font-size:1.8rem;margin-bottom:16px;border-bottom:2px solid var(--border);padding-bottom:8px}}
+.content h2{{font-size:1.4rem;margin:24px 0 12px}}
+.content h3{{font-size:1.15rem;margin:20px 0 8px}}
+.content p{{margin:8px 0}}
+.content pre{{background:var(--bg-card);border:1px solid var(--border);border-radius:8px;padding:12px 16px;overflow-x:auto;font-size:0.85rem}}
+.content code{{font-size:0.9em;background:var(--bg-card);padding:1px 4px;border-radius:3px}}
+.content pre code{{background:none;padding:0}}
+.content a{{color:var(--accent)}}.content img{{max-width:100%;border-radius:6px}}
+.fm-section{{background:var(--bg-card);border:1px solid var(--border);border-radius:8px;padding:12px 16px;margin-bottom:20px}}
+.fm-header{{font-weight:600;font-size:0.8rem;color:var(--text-muted);text-transform:uppercase;letter-spacing:0.5px;margin-bottom:8px}}
+.fm-body{{display:grid;grid-template-columns:1fr 2fr;gap:4px 12px;font-size:0.85rem}}
+.fm-row{{display:contents}}
+.fm-key{{color:var(--accent);font-weight:500}}
+.fm-val{{color:var(--text);word-break:break-word}}
+.content blockquote{{border-left:3px solid var(--accent);padding-left:16px;color:var(--text-muted);margin:12px 0}}
+.content table{{border-collapse:collapse;width:100%;margin:12px 0}}
+.content th,.content td{{border:1px solid var(--border);padding:8px 12px;text-align:left}}
+.content th{{background:var(--bg-card)}}
+@media print{{.toolbar,.share-banner{{display:none}}body{{background:#fff;color:#000}}}}
+@media(max-width:600px){{.content{{padding:16px 12px 40px}}.toolbar{{gap:4px}}.toolbar-btn{{padding:4px 8px;font-size:0.7rem}}}}
+</style></head>
+<body>
+<div class="share-banner">
+  <svg xmlns="http://www.w3.org/2000/svg" width="14" height="14" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"><path d="M14.5 2H6a2 2 0 0 0-2 2v16a2 2 0 0 0 2 2h12a2 2 0 0 0 2-2V7.5L14.5 2z"/><polyline points="14 2 14 8 20 8"/></svg>
+  Document partagé via ObsiGate
+</div>
+<div class="toolbar">
+  <span class="toolbar-title">{title_esc}</span>
+  <button class="toolbar-btn" data-share-theme title="Thème clair/sombre">
+    <svg id="theme-icon-dark" xmlns="http://www.w3.org/2000/svg" width="15" height="15" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"><path d="M21 12.79A9 9 0 1 1 11.21 3 7 7 0 0 0 21 12.79z"/></svg>
+    <svg id="theme-icon-light" xmlns="http://www.w3.org/2000/svg" width="15" height="15" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" style="display:none"><circle cx="12" cy="12" r="5"/><line x1="12" y1="1" x2="12" y2="3"/><line x1="12" y1="21" x2="12" y2="23"/><line x1="4.22" y1="4.22" x2="5.64" y2="5.64"/><line x1="18.36" y1="18.36" x2="19.78" y2="19.78"/><line x1="1" y1="12" x2="3" y2="12"/><line x1="21" y1="12" x2="23" y2="12"/><line x1="4.22" y1="19.78" x2="5.64" y2="18.36"/><line x1="18.36" y1="5.64" x2="19.78" y2="4.22"/></svg>
+  </button>
+  <button class="toolbar-btn" data-share-md title="Télécharger en Markdown">
+    <svg xmlns="http://www.w3.org/2000/svg" width="15" height="15" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"><path d="M21 15v4a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2v-4"/><polyline points="7 10 12 15 17 10"/><line x1="12" y1="15" x2="12" y2="3"/></svg>
+    .md
+  </button>
+  <button class="toolbar-btn" data-share-pdf title="Télécharger en PDF">
+    <svg xmlns="http://www.w3.org/2000/svg" width="15" height="15" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"><path d="M14 2H6a2 2 0 0 0-2 2v16a2 2 0 0 0 2-2V8z"/><polyline points="14 2 14 8 20 8"/><line x1="16" y1="13" x2="8" y2="13"/><line x1="16" y1="17" x2="8" y2="17"/><polyline points="10 9 9 9 8 9"/></svg>
+    PDF
+  </button>
+</div>
+<div class="content" id="content">{fm_html}{html}</div>
+<script id="raw-content" type="text/plain" style="display:none">{raw_json}</script>
+<script>
+function toggleTheme(){{var t=document.documentElement;var isDark=t.dataset.theme==="dark";t.dataset.theme=isDark?"light":"dark";document.getElementById("theme-icon-dark").style.display=isDark?"none":"";document.getElementById("theme-icon-light").style.display=isDark?"":"none";localStorage.setItem("obsigate-share-theme",t.dataset.theme)}}
+(function(){{var s=localStorage.getItem("obsigate-share-theme");if(!s)s="dark";document.documentElement.dataset.theme=s;var isDark=s==="dark";document.getElementById("theme-icon-dark").style.display=isDark?"":"none";document.getElementById("theme-icon-light").style.display=isDark?"none":""}})();
+function exportMD(){{var raw=JSON.parse(document.getElementById("raw-content").textContent);var b=new Blob([raw],{{type:"text/markdown"}});var a=document.createElement("a");a.href=URL.createObjectURL(b);a.download={title_download_js};a.click()}}
+document.querySelector("[data-share-theme]").addEventListener("click",toggleTheme);
+document.querySelector("[data-share-md]").addEventListener("click",exportMD);
+document.querySelector("[data-share-pdf]").addEventListener("click",function(){{location.href=location.pathname+"/pdf"}});
+</script></body></html>""",
+            request.state.csp_nonce,
+        ),
+    )
@@ -0,0 +1,107 @@
+"""Vault management endpoints (ROADMAP #85, tranche 8).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/vaults*``), mêmes modèles de réponse
+(``VaultInfo`` déménagé dans :mod:`backend.schemas`), mêmes dépendances
+d'authentification.
+
+Le handle du file-watcher vit désormais dans :mod:`backend.watcher_state`
+(partagé avec le lifespan de ``main``) au lieu du global de ``main``.
+"""
+
+from pathlib import Path
+
+from fastapi import APIRouter, Body, Depends, HTTPException
+
+from backend.auth.middleware import require_admin, require_auth
+from backend.indexer import add_vault_to_index, index, remove_vault_from_index
+from backend.schemas import VaultActionResponse, VaultInfo, VaultsStatusResponse, VaultStatsResponse
+from backend.services.vaults import list_accessible_vaults
+from backend.sse import sse_manager
+from backend.watcher_state import get_watcher
+
+router = APIRouter(tags=["vaults"])
+
+
+@router.get("/api/vaults", response_model=list[VaultInfo])
+async def api_vaults(current_user=Depends(require_auth)):
+    """List configured vaults the user has access to.
+
+    Returns:
+        List of vault summary objects filtered by user permissions.
+    """
+    return list_accessible_vaults(current_user)
+
+
+@router.post("/api/vaults/add", response_model=VaultStatsResponse)
+async def api_add_vault(body: dict = Body(...), current_user=Depends(require_admin)):
+    """Add a new vault dynamically without restarting.
+
+    Body:
+        name: Display name for the vault.
+        path: Absolute filesystem path to the vault directory.
+    """
+    name = body.get("name", "").strip()
+    vault_path = body.get("path", "").strip()
+
+    if not name or not vault_path:
+        raise HTTPException(status_code=400, detail="Both 'name' and 'path' are required")
+
+    if name in index:
+        raise HTTPException(status_code=409, detail=f"Vault '{name}' already exists")
+
+    if not Path(vault_path).exists():
+        raise HTTPException(status_code=400, detail=f"Path does not exist: {vault_path}")
+
+    stats = await add_vault_to_index(name, vault_path)
+
+    # Start watching the new vault
+    watcher = get_watcher()
+    if watcher:
+        await watcher.add_vault(name, vault_path)
+
+    await sse_manager.broadcast("vault_added", {"vault": name, "stats": stats})
+    return {"status": "ok", "vault": name, "stats": stats}
+
+
+@router.delete("/api/vaults/{vault_name}", response_model=VaultActionResponse)
+async def api_remove_vault(vault_name: str, current_user=Depends(require_admin)):
+    """Remove a vault from the index and stop watching it.
+
+    Args:
+        vault_name: Name of the vault to remove.
+    """
+    if vault_name not in index:
+        raise HTTPException(status_code=404, detail=f"Vault '{vault_name}' not found")
+
+    # Stop watching
+    watcher = get_watcher()
+    if watcher:
+        await watcher.remove_vault(vault_name)
+
+    await remove_vault_from_index(vault_name)
+    await sse_manager.broadcast("vault_removed", {"vault": vault_name})
+    return {"status": "ok", "vault": vault_name}
+
+
+@router.get("/api/vaults/status", response_model=VaultsStatusResponse)
+async def api_vaults_status(current_user=Depends(require_auth)):
+    """Detailed status of all vaults including watcher state.
+
+    Returns per-vault: file count, tag count, watching status, vault path.
+    """
+    watcher = get_watcher()
+    statuses = {}
+    for vname, vdata in index.items():
+        watching = watcher is not None and vname in watcher.observers
+        statuses[vname] = {
+            "file_count": len(vdata.get("files", [])),
+            "tag_count": len(vdata.get("tags", {})),
+            "path": vdata.get("path", ""),
+            "watching": watching,
+        }
+    return {
+        "vaults": statuses,
+        "watcher_active": watcher is not None,
+        "sse_clients": sse_manager.client_count,
+    }
@@ -0,0 +1,54 @@
+"""Webhook CRUD endpoints (ROADMAP #85, tranche 2).
+
+Handlers déplacés depuis :mod:`backend.main` sans changement de
+comportement : mêmes chemins (``/api/webhooks``), même modèle de réponse
+(:class:`backend.schemas.WebhookModel`), même dépendance admin. La logique
+métier vit déjà dans :mod:`backend.webhooks` (validation d'URL anti-SSRF,
+store ``webhook_secrets.json`` — BUG-026).
+"""
+
+from fastapi import APIRouter, Body, Depends, HTTPException
+
+from backend.auth.middleware import require_admin
+from backend.schemas import StatusResponse, WebhookModel
+from backend.webhooks import (
+    create_webhook,
+    delete_webhook,
+    get_webhooks,
+    update_webhook,
+)
+
+router = APIRouter(prefix="/api/webhooks", tags=["webhooks"])
+
+
+@router.get("", response_model=list[WebhookModel])
+async def api_webhooks_list(current_user=Depends(require_admin)):
+    return get_webhooks()
+
+
+@router.post("", response_model=WebhookModel)
+async def api_webhooks_create(body: dict = Body(...), current_user=Depends(require_admin)):
+    name = body.get("name", "Unnamed")
+    url = body.get("url", "")
+    events = body.get("events", [])
+    secret = body.get("secret")
+    if not url:
+        raise HTTPException(400, "URL is required")
+    return create_webhook(name, url, events, secret)
+
+
+@router.patch("/{webhook_id}", response_model=WebhookModel)
+async def api_webhooks_update(
+    webhook_id: str, body: dict = Body(...), current_user=Depends(require_admin)
+):
+    result = update_webhook(webhook_id, body)
+    if not result:
+        raise HTTPException(404, "Webhook not found")
+    return result
+
+
+@router.delete("/{webhook_id}", response_model=StatusResponse)
+async def api_webhooks_delete(webhook_id: str, current_user=Depends(require_admin)):
+    if not delete_webhook(webhook_id):
+        raise HTTPException(404, "Webhook not found")
+    return {"status": "deleted"}
@@ -188,6 +188,486 @@ class BackupsAutoResponse(BaseModel):
    since_hours: int | float = Field(description="Look-back window in hours")


+class DiffResponse(BaseModel):
+    """Response containing a unified diff between two file versions (#85 — extrait de backend.main, inchangé)."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path")
+    version: int = Field(description="Backup version timestamp (left/old side)")
+    compare_with: int | None = Field(default=None, description="Other backup version or null for current file (right/new side)")
+    diff: str = Field(description="Unified diff (empty if no changes)")
+
+
+class RestoreRequest(BaseModel):
+    """Request to restore a file from a backup (#85 — extrait de backend.main, inchangé)."""
+
+    version: int = Field(description="Timestamp of the backup version to restore")
+
+
+class RestoreResponse(BaseModel):
+    """Response after restoring a file from backup (#85 — extrait de backend.main, inchangé)."""
+
+    success: bool = Field(description="Whether restore succeeded")
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path")
+    restored_from: int = Field(description="Timestamp of the backup used")
+    current_backed_up: int | None = Field(default=None, description="Timestamp of the backup created from the current version before restore, if any")
+
+
+class BackupEntry(BaseModel):
+    """A single backup version of a file (#85 — extrait de backend.main, inchangé)."""
+
+    timestamp: int = Field(description="Unix timestamp of when the backup was created")
+    datetime: str = Field(description="ISO 8601 datetime string")
+    size: int = Field(description="File size in bytes")
+    filename: str = Field(description="Backup filename on disk")
+
+
+class BackupListResponse(BaseModel):
+    """Response listing all available backups for a file (#85 — extrait de backend.main, inchangé)."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path")
+    backups: list[BackupEntry] = Field(description="Available backups, newest first")
+
+
+class DiffRequest(BaseModel):
+    """Request parameters for generating a diff (#85 — extrait de backend.main, inchangé)."""
+
+    version: int = Field(description="Timestamp of the backup version to compare")
+    compare_with: int | None = Field(default=None, description="Timestamp of another backup version. If omitted, compares with the current file.")
+
+
+# ---------------------------------------------------------------------------
+# Files — browse / read (#85 — extrait de backend.main, inchangé)
+# ---------------------------------------------------------------------------
+
+
+class BrowseItem(BaseModel):
+    """A single entry (file or directory) returned by the browse endpoint."""
+
+    name: str = Field(description="File or directory name")
+    path: str = Field(description="Relative path within vault")
+    type: str = Field(description="'file' or 'directory'")
+    children_count: int | None = Field(default=None, description="Number of children (directories only)")
+    size: int | None = Field(default=None, description="File size in bytes")
+    extension: str | None = Field(default=None, description="File extension")
+
+
+class BrowseResponse(BaseModel):
+    """Paginated directory listing for a vault."""
+
+    vault: str
+    path: str
+    items: list[BrowseItem]
+
+
+class FileContentResponse(BaseModel):
+    """Rendered file content with metadata."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path within the vault")
+    title: str = Field(description="File title (from frontmatter or filename)")
+    tags: list[str] = Field(description="Extracted tags from frontmatter and inline #tags")
+    frontmatter: dict[str, Any] = Field(description="YAML frontmatter as key-value dict")
+    html: str = Field(description="Rendered HTML content")
+    raw_length: int = Field(description="Length of raw file content in characters")
+    extension: str = Field(description="File extension (e.g. .md, .txt)")
+    is_markdown: bool = Field(description="Whether the file is markdown")
+    unsupported: bool | None = Field(default=False, description="True for binary/unsupported files")
+    size_bytes: int | None = Field(default=None, description="File size in bytes (for unsupported files)")
+    is_pdf: bool | None = Field(default=None, description="True for PDF files")
+    is_image: bool | None = Field(default=None, description="True for image files")
+    is_audio: bool | None = Field(default=None, description="True for audio files (HTML5 <audio>, roadmap #109)")
+    is_video: bool | None = Field(default=None, description="True for video files (HTML5 <video>, roadmap #109)")
+    media_too_large: bool | None = Field(default=None, description="True when audio/video exceeds the inline streaming limit")
+    stream_url: str | None = Field(default=None, description="Byte-range streaming URL under /api/media (audio/video)")
+    media_mime: str | None = Field(default=None, description="MIME type for audio/video files")
+    is_csv: bool | None = Field(default=None, description="True for CSV files")
+    is_xlsx: bool | None = Field(default=None, description="True for Excel .xlsx files")
+    xlsx_sheets: list[dict[str, Any]] | None = Field(
+        default=None,
+        description=(
+            "Rendered xlsx sheets [{name, html, rows, cols, total_rows, "
+            "total_cols, max_rows, max_cols, truncated}] — `truncated` is true "
+            "when the sheet exceeds the 500x40 render caps (#153 A8)"
+        ),
+    )
+    xlsx_lossy_features: list[str] | None = Field(
+        default=None,
+        description=(
+            "Workbook parts an openpyxl save would drop (#153 A1) — e.g. "
+            "cached_values, slicers, form_controls, connections, custom_xml, "
+            "signature, rich_comments, macros. Empty/absent = nothing at risk."
+        ),
+    )
+    is_json: bool | None = Field(default=None, description="True for JSON files")
+    is_excalidraw: bool | None = Field(default=None, description="True for Excalidraw diagram files")
+    excalidraw_data: dict[str, Any] | None = Field(default=None, description="Excalidraw diagram data (elements, appState, files)")
+    excalidraw_data_compressed: str | None = Field(default=None, description="Compressed Excalidraw data for .excalidraw.md files")
+    pdf_metadata: dict[str, Any] | None = Field(default=None, description="PDF metadata")
+    pdf_toc: list[dict[str, Any]] | None = Field(default=None, description="PDF table of contents")
+    image_mime: str | None = Field(default=None, description="MIME type for image files")
+
+
+class XlsxSheetWindowResponse(BaseModel):
+    """One window of rows of a single .xlsx sheet (lazy loading, #153 A9).
+
+    Served by ``GET /api/file/{vault_name}/xlsx/sheet``; the row numbers and
+    the ``data-cell`` references in ``html`` are the real A1 coordinates of the
+    sheet, whatever the window.
+    """
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path within the vault")
+    sheet: str = Field(description="Sheet name (as shown in the tab)")
+    offset: int = Field(description="0-based index of the first returned row")
+    limit: int = Field(description="Maximum number of rows returned (capped server-side)")
+    rows: int = Field(description="Rows actually returned in this window")
+    cols: int = Field(description="Columns of the rendered window")
+    total_rows: int = Field(description="Rows the sheet declares")
+    total_cols: int = Field(description="Columns the sheet declares")
+    max_rows: int = Field(description="Row cap of the renderer (500) — the coverage of this window")
+    max_cols: int = Field(description="Column cap of the renderer (40)")
+    truncated: bool = Field(
+        description="True when the sheet exceeds the 500x40 render caps"
+    )
+    has_more: bool = Field(description="True when rows remain after this window")
+    html: str = Field(description="Rendered HTML table for the window")
+
+
+class FileRawResponse(BaseModel):
+    """Raw text content of a file."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path within the vault")
+    raw: str = Field(description="Raw file content as text")
+
+
+# ---------------------------------------------------------------------------
+# Files — mutations (#85 — extrait de backend.main, inchangé)
+# ---------------------------------------------------------------------------
+
+
+class FileSaveResponse(BaseModel):
+    """Confirmation after saving a file."""
+
+    status: str = Field(description="Always 'ok'")
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path within the vault")
+    size: int = Field(description="Size of saved content in characters")
+
+
+class FileDeleteResponse(BaseModel):
+    """Confirmation after deleting a file."""
+
+    status: str = Field(description="Always 'ok'")
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path within the vault")
+
+
+class DirectoryCreateRequest(BaseModel):
+    """Request to create a new directory."""
+
+    path: str = Field(description="Relative path of the new directory")
+
+
+class DirectoryCreateResponse(BaseModel):
+    """Response after creating a directory."""
+
+    success: bool = Field(description="Whether creation succeeded")
+    path: str = Field(description="Path of the created directory")
+
+
+class DirectoryRenameRequest(BaseModel):
+    """Request to rename a directory."""
+
+    path: str = Field(description="Current path of the directory")
+    new_name: str = Field(description="New name for the directory")
+
+
+class DirectoryRenameResponse(BaseModel):
+    """Response after renaming a directory."""
+
+    success: bool = Field(description="Whether rename succeeded")
+    old_path: str = Field(description="Original directory path")
+    new_path: str = Field(description="New directory path")
+
+
+class DirectoryDeleteResponse(BaseModel):
+    """Response after deleting a directory."""
+
+    success: bool = Field(description="Whether deletion succeeded")
+    deleted_count: int = Field(description="Number of files recursively deleted")
+
+
+class FileCreateRequest(BaseModel):
+    """Request to create a new file."""
+
+    path: str = Field(description="Relative path of the new file")
+    content: str = Field(default="", description="Initial content")
+
+
+class FileCreateResponse(BaseModel):
+    """Response after creating a file."""
+
+    success: bool = Field(description="Whether creation succeeded")
+    path: str = Field(description="Path of the created file")
+
+
+class BatchUploadFileItem(BaseModel):
+    """A single file/dir entry in a batch upload request."""
+
+    path: str = Field(description="Relative path of the item within the batch")
+    content: str | None = Field(default=None, description="Base64 encoded or text content for files")
+    is_dir: bool = Field(default=False, description="True if entry represents an empty directory")
+
+
+class BatchUploadRequest(BaseModel):
+    """Request payload for batch file/directory upload."""
+
+    target_dir: str = Field(default="", description="Base directory in vault to upload into (empty for root)")
+    files: list[BatchUploadFileItem] = Field(description="List of files and directories to upload")
+    overwrite: bool = Field(default=True, description="Whether to overwrite existing files (creates backups)")
+
+
+class BatchUploadResponse(BaseModel):
+    """Response from batch file/directory upload."""
+
+    success: bool = Field(description="True if all files uploaded without error")
+    vault: str = Field(description="Vault name")
+    target_dir: str = Field(description="Target directory")
+    uploaded: list[str] = Field(description="List of created/updated file paths")
+    created_dirs: list[str] = Field(description="List of created directory paths")
+    errors: list[dict[str, Any]] = Field(default_factory=list, description="List of items that failed")
+    total_files: int = Field(description="Total uploaded files count")
+
+
+class FileRenameRequest(BaseModel):
+    """Request to rename a file."""
+
+    path: str = Field(description="Current path of the file")
+    new_name: str = Field(description="New name for the file")
+
+
+class FileRenameResponse(BaseModel):
+    """Response after renaming a file."""
+
+    success: bool = Field(description="Whether rename succeeded")
+    old_path: str
+    new_path: str
+
+
+class FileMoveRequest(BaseModel):
+    """Request to move a file or directory to a different parent directory."""
+
+    source_path: str = Field(description="Current relative path of the file/directory")
+    destination_dir: str = Field(description="Target directory relative path (empty string for vault root)")
+
+
+class FileMoveResponse(BaseModel):
+    """Response after moving a file or directory."""
+
+    success: bool = Field(description="Whether move succeeded")
+    old_path: str = Field(description="Original path")
+    new_path: str = Field(description="New path after move")
+    item_type: str = Field(description="Type of item moved: 'file' or 'directory'")
+
+
+# ---------------------------------------------------------------------------
+# Vaults & history (#85 — extrait de backend.main, inchangé)
+# ---------------------------------------------------------------------------
+
+
+class VaultInfo(BaseModel):
+    """Summary information about a configured vault."""
+
+    name: str = Field(description="Display name of the vault")
+    file_count: int = Field(description="Number of indexed files")
+    tag_count: int = Field(description="Number of unique tags")
+    type: str = Field(default="VAULT", description="Type of the vault mapping (VAULT or DIR)")
+
+
+class BookmarkToggleRequest(BaseModel):
+    """Request to toggle a bookmark on a file."""
+
+    vault: str
+    path: str
+    title: str | None = None
+
+
+# ---------------------------------------------------------------------------
+# Search / suggest / graph (#85 — extrait de backend.main, inchangé)
+# ---------------------------------------------------------------------------
+
+
+class SearchResultItem(BaseModel):
+    """A single search result."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path")
+    title: str = Field(description="File title")
+    tags: list[str] = Field(description="File tags")
+    score: int = Field(description="Relevance score")
+    snippet: str = Field(description="Content excerpt with highlights")
+    modified: str = Field(description="ISO 8601 modification timestamp")
+
+
+class SearchResponse(BaseModel):
+    """Full-text search response with optional pagination."""
+
+    query: str = Field(description="Original search query")
+    vault_filter: str = Field(description="Vault filter applied ('all' or vault name)")
+    tag_filter: str | None = Field(default=None, description="Tag filter applied")
+    count: int = Field(description="Number of results in this response")
+    total: int = Field(default=0, description="Total results before pagination")
+    offset: int = Field(default=0, description="Current pagination offset")
+    limit: int = Field(default=200, description="Page size")
+    results: list[SearchResultItem] = Field(description="Search result items")
+
+
+class TagsResponse(BaseModel):
+    """Tag aggregation response."""
+
+    vault_filter: str | None = Field(default=None, description="Vault filter applied")
+    tags: dict[str, int] = Field(description="Tag name → count mapping")
+
+
+class TreeSearchResult(BaseModel):
+    """A single tree search result item."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Full relative path")
+    name: str = Field(description="File or directory name")
+    type: str = Field(description="'file' or 'directory'")
+    matched_path: str = Field(description="Path segment that matched the query")
+
+
+class TreeSearchResponse(BaseModel):
+    """Tree search response with matching paths."""
+
+    query: str = Field(description="Search query")
+    vault_filter: str = Field(description="Vault filter applied")
+    results: list[TreeSearchResult] = Field(description="Matching files and directories")
+
+
+class VaultPathEntry(BaseModel):
+    """A single indexed path (file or directory) in a vault."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Full relative path")
+    name: str = Field(description="File or directory name")
+    type: str = Field(description="'file' or 'directory'")
+
+
+class VaultPathsResponse(BaseModel):
+    """Flat list of every indexed path in a vault (capped)."""
+
+    vault: str = Field(description="Vault name")
+    count: int = Field(description="Number of returned entries")
+    results: list[VaultPathEntry] = Field(description="Indexed files and directories")
+
+
+class AdvancedSearchResultItem(BaseModel):
+    """A single advanced search result with highlighted snippet."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path")
+    title: str = Field(description="File title")
+    tags: list[str] = Field(description="File tags")
+    score: float = Field(description="TF-IDF relevance score (or fused RRF score in semantic mode)")
+    semantic_score: float = Field(default=0.0, description="Cosine similarity from the semantic index (0 when unavailable)")
+    snippet: str = Field(description="Content excerpt with <mark> highlights")
+    modified: str = Field(description="ISO 8601 modification timestamp")
+    extension: str = Field(default="", description="File extension")
+
+
+class SearchFacets(BaseModel):
+    """Faceted counts for search results."""
+
+    tags: dict[str, int] = Field(default_factory=dict)
+    vaults: dict[str, int] = Field(default_factory=dict)
+
+
+class AdvancedSearchResponse(BaseModel):
+    """Advanced search response with TF-IDF scoring, facets, and pagination."""
+
+    results: list[AdvancedSearchResultItem] = Field(description="Search results")
+    total: int = Field(description="Total number of matching results")
+    offset: int = Field(description="Current pagination offset")
+    limit: int = Field(description="Page size")
+    facets: SearchFacets = Field(description="Faceted counts by tag and vault")
+    query_time_ms: float = Field(default=0, description="Server-side query time in milliseconds")
+    semantic_available: bool = Field(default=False, description="True when the semantic (embedding) index is ready")
+
+
+class TitleSuggestion(BaseModel):
+    """A file title suggestion for autocomplete."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Relative file path")
+    title: str = Field(description="File title")
+
+
+class SuggestResponse(BaseModel):
+    """Autocomplete suggestions for file titles."""
+
+    query: str = Field(description="Original query string")
+    suggestions: list[TitleSuggestion] = Field(description="Matching file suggestions")
+
+
+class TagSuggestion(BaseModel):
+    """A tag suggestion for autocomplete."""
+
+    tag: str = Field(description="Tag name")
+    count: int = Field(description="Number of files with this tag")
+
+
+class TagSuggestResponse(BaseModel):
+    """Autocomplete suggestions for tags."""
+
+    query: str = Field(description="Original query string")
+    suggestions: list[TagSuggestion] = Field(description="Matching tag suggestions")
+
+
+class GraphNode(BaseModel):
+    """A single node in the graph view."""
+
+    id: str = Field(description="Unique node identifier")
+    name: str = Field(description="Display name")
+    type: str = Field(description="'vault', 'directory', or 'file'")
+    path: str = Field(description="Relative path within vault")
+    size: int = Field(default=0, description="File size in bytes")
+    tags: list[str] = Field(default_factory=list, description="Tags from frontmatter")
+    incoming_count: int = Field(default=0, description="Number of incoming wikilinks")
+    outgoing_count: int = Field(default=0, description="Number of outgoing wikilinks")
+
+
+class GraphEdge(BaseModel):
+    """An edge between two nodes in the graph view."""
+
+    source: str = Field(description="Source node ID")
+    target: str = Field(description="Target node ID")
+    relation: str = Field(description="'parent', 'wikilink', or 'backlink'")
+
+
+class GraphResponse(BaseModel):
+    """Graph data for a vault or directory."""
+
+    vault: str = Field(description="Vault name")
+    path: str = Field(description="Root path for the graph")
+    scope: str = Field(default="directory", description="'directory' or 'full'")
+    nodes: list[GraphNode] = Field(description="Graph nodes (files and directories)")
+    edges: list[GraphEdge] = Field(description="Graph edges (parent and wikilink relations)")
+
+
+class ReloadResponse(BaseModel):
+    """Index reload confirmation with per-vault stats."""
+
+    status: str = Field(description="Reload status ('ok' or 'error')")
+    vaults: dict[str, Any] = Field(description="Per-vault file counts after reload")
+
+
 # ---------------------------------------------------------------------------
 # PDF
 # ---------------------------------------------------------------------------
@@ -408,6 +888,31 @@ class DashboardResponse(BaseModel):
    total_images: int = 0


+# ---------------------------------------------------------------------------
+# System / health (#85 — extrait de backend.main, comportement inchangé)
+# ---------------------------------------------------------------------------
+
+
+class HealthResponse(BaseModel):
+    """Application health status.
+
+    Déplacé depuis :mod:`backend.main` sans modification : pas de
+    ``extra="allow"`` ici, pour préserver la validation actuelle des
+    réponses (les champs enrichis de ``/api/health/detailed`` restent
+    filtrés comme avant).
+    """
+
+    status: str = Field(description="Health status ('ok' or 'error')")
+    version: str = Field(description="Application version (x.y.z — latest release tag)")
+    vaults: int = Field(description="Number of configured vaults")
+    total_files: int = Field(description="Total indexed files across all vaults")
+    total_tokens: int = Field(description="Total indexed tokens (approx.) across all vaults", default=0)
+    last_full_index_ts: str = Field(description="ISO timestamp of last full index rebuild", default="")
+    uptime_seconds: int = Field(description="Server uptime in seconds", default=0)
+    git_describe: str = Field(default="", description="Full git describe string (commits beyond tag), empty if no git")
+    git_commit: str = Field(default="", description="Short HEAD commit hash, empty if no git")
+
+
 # ---------------------------------------------------------------------------
 # Webhooks, sharing & conflicts
 # ---------------------------------------------------------------------------
@@ -12,7 +12,12 @@ from sortedcontainers import SortedList

 from backend import indexer as _indexer
 from backend import semantic_search as _semantic
-from backend.indexer import index
+
+# NOTE: the shared index is read through ``_indexer.index`` everywhere, never
+# via ``from backend.indexer import index``. That import binds the dict object
+# once, so a module reload of ``backend.indexer`` (tests, dev reload) rebinds
+# the module-level name to a FRESH dict while this module keeps writing to the
+# stale one — the inverted index then silently indexes nothing (BUG-089).
 from backend.services.regex_safety import (
    MAX_REGEX_MATCHES,
    truncate_for_regex,
@@ -371,9 +376,15 @@ class InvertedIndex:
        self._sorted_tokens: SortedList = SortedList()
        self._ready: bool = False  # True after initial build

-    def is_stale(self) -> bool:
-        """Return True if the index has not been built yet."""
-        return not self._ready
+    def is_ready(self) -> bool:
+        """Return True once the initial build has completed.
+
+        The index is then kept current incrementally by ``add_document()`` /
+        ``remove_document()``, so it never goes stale: there is no generation
+        counter, no cooldown and no lazy rebuild. Searches simply fall back to
+        a full scan while this is False (see ``search()``).
+        """
+        return self._ready

    def rebuild(self) -> None:
        """Rebuild inverted index from the global ``index`` dict.
@@ -393,7 +404,7 @@ class InvertedIndex:
        self.vault_docs = defaultdict(set)
        self.tag_docs = defaultdict(set)

-        for vault_name, vault_data in index.items():
+        for vault_name, vault_data in _indexer.index.items():
            for file_info in vault_data.get("files", []):
                doc_key = f"{vault_name}::{file_info['path']}"
                self.doc_count += 1
@@ -537,6 +548,10 @@ class InvertedIndex:
        self.doc_vault.pop(doc_key, None)
        if vault_name in self.vault_docs:
            self.vault_docs[vault_name].discard(doc_key)
+            # Drop the empty entry so a fully removed vault leaves no trace
+            # (it is a defaultdict: a bare lookup would recreate the key).
+            if not self.vault_docs[vault_name]:
+                del self.vault_docs[vault_name]
        # Tags (per-document, NOT the global tag_norm_map)
        for tag in file_info.get("tags", []):
            td = self.tag_docs.get(tag.lower())
@@ -678,7 +693,7 @@ _indexer.set_index_change_hook(_on_index_change_hook)

 def init_inverted_index():
    """Force initial inverted index build. Called after build_index completes on startup."""
-    if any(vdata.get("files") for vdata in index.values()):
+    if any(vdata.get("files") for vdata in _indexer.index.values()):
        _inverted_index.rebuild()
        logger.info("Inverted index initialized.")

@@ -739,7 +754,7 @@ def search(
    results: list[dict[str, Any]] = []

    inv = get_inverted_index()
-    use_index = (not inv.is_stale()) and inv.doc_count > 0
+    use_index = inv.is_ready() and inv.doc_count > 0

    if use_index:
        # BUG-033: retrieve candidates from the inverted index instead of
@@ -774,7 +789,7 @@ def search(
    else:
        candidates = [
            (vault_name, file_info)
-            for vault_name, vault_data in index.items()
+            for vault_name, vault_data in _indexer.index.items()
            if vault_filter == "all" or vault_name == vault_filter
            for file_info in vault_data["files"]
        ]
@@ -1603,7 +1618,7 @@ def get_all_tags(vault_filter: str | None = None) -> dict[str, int]:
        Dict mapping tag names to their total occurrence count.
    """
    merged: dict[str, int] = {}
-    for vault_name, vault_data in index.items():
+    for vault_name, vault_data in _indexer.index.items():
        if vault_filter and vault_filter != "all" and vault_name != vault_filter:
            continue
        for tag, count in vault_data.get("tags", {}).items():
@@ -0,0 +1,35 @@
+"""Shared thread pool for CPU-bound search (ROADMAP #85, tranche 5).
+
+Holder extrait de :mod:`backend.main` sans changement de comportement :
+un seul pool (2 workers, préfixe ``"search"``) créé au démarrage et arrêté
+à l'extinction par le lifespan de ``main``. Les routers et les endpoints
+restants y accèdent via :func:`get_search_executor` au lieu du global de
+``main`` (plus d'import circulaire potentiel).
+"""
+
+from __future__ import annotations
+
+from concurrent.futures import ThreadPoolExecutor
+
+_executor: ThreadPoolExecutor | None = None
+
+
+def init_search_executor(max_workers: int = 2) -> ThreadPoolExecutor:
+    """Create (or reuse) the shared search thread pool."""
+    global _executor
+    if _executor is None:
+        _executor = ThreadPoolExecutor(max_workers=max_workers, thread_name_prefix="search")
+    return _executor
+
+
+def shutdown_search_executor() -> None:
+    """Stop the shared search thread pool (best-effort, non-blocking)."""
+    global _executor
+    if _executor is not None:
+        _executor.shutdown(wait=False)
+        _executor = None
+
+
+def get_search_executor() -> ThreadPoolExecutor | None:
+    """Return the shared search thread pool (``None`` before startup)."""
+    return _executor
@@ -457,10 +457,6 @@ class SemanticIndex:
        """Return True once a full rebuild has completed."""
        return self._ready

-    def is_stale(self) -> bool:
-        """Alias used by callers that check index freshness."""
-        return not self._ready
-
    def _ensure_provider(self) -> EmbeddingProvider:
        if self.provider is None:
            self.provider = get_embedding_provider()
@@ -31,7 +31,7 @@ DEFAULT_MAX_BACKUPS = 10
 def _default_max_backups() -> int:
    """Read ``max_backups_per_file`` from app config (lazy, best-effort)."""
    try:
-        from backend.main import _load_config
+        from backend.routers.config import _load_config  # ROADMAP #85 T7 — déménagé depuis backend.main

        return int(_load_config().get("max_backups_per_file", DEFAULT_MAX_BACKUPS))
    except Exception:  # pragma: no cover - config unavailable
@@ -14,8 +14,12 @@ from __future__ import annotations

 import logging
 import os
+import re
 import shutil
-from collections.abc import Callable
+import threading
+from collections.abc import Callable, Iterator
+from contextlib import contextmanager
+from datetime import date, datetime
 from pathlib import Path
 from typing import Any

@@ -223,6 +227,234 @@ def edit_file(
    return {"success": True, "vault": vault_name, "path": rel_path, "size": len(content)}


+# Cell reference like "A1" / "AB42" (Excel A1 notation, up to 3 letters / 8 digits).
+_XLSX_CELL_RE = re.compile(r"^[A-Z]{1,3}[1-9][0-9]{0,7}$")
+# ponytail: bare int/float coercion mirrors what Excel does when you type a
+# number; dates/booleans stay text (upgrade path: parse locale dates too).
+_XLSX_INT_RE = re.compile(r"^[+-]?\d+$")
+_XLSX_FLOAT_RE = re.compile(r"^[+-]?(?:\d+\.\d*|\.\d+)$")
+# #153 A4 — openpyxl turns any string starting with "=" into a formula, which
+# Excel then evaluates on open (DDE / =cmd|… / =HYPERLINK exfiltration). "@" is
+# the legacy Lotus-style trigger. "+"/"-" are left alone: they are numbers here.
+_XLSX_FORMULA_RE = re.compile(r"^[=@]")
+
+# #153 A10 — types recognised when a user types into a cell. Excel infers them
+# too; storing everything as text would make a spreadsheet unusable (a boolean
+# column stays a string, a date column sorts lexicographically).
+_XLSX_TRUE_LITERALS = {"true", "vrai", "oui", "yes"}
+_XLSX_FALSE_LITERALS = {"false", "faux", "non", "no"}
+# Shape check before strptime: keeps the hot path free of format attempts.
+_XLSX_DATE_RE = re.compile(r"^\d{1,2}[-/]\d{1,2}[-/]\d{4}(?:[ T]\d{1,2}:\d{2})?$")
+
+# #153 A3 — per-file write lock. Two concurrent saves (two tabs, the AI agent
+# and the viewer, a watcher restore) would otherwise read-modify-write on the
+# same archive and the last writer silently wins. Kept deliberately small: the
+# lock only covers the load → edit → atomic-replace window.
+_XLSX_LOCK_TIMEOUT = 15.0
+_xlsx_locks: dict[str, threading.Lock] = {}
+_xlsx_locks_guard = threading.Lock()
+
+
+@contextmanager
+def _xlsx_write_lock(key: str) -> Iterator[None]:
+    """Serialize the read-modify-write of one workbook path.
+
+    Raises:
+        ServiceError: ``conflict`` (409) when the lock is still held after
+            :data:`_XLSX_LOCK_TIMEOUT` seconds.
+    """
+    with _xlsx_locks_guard:
+        lock = _xlsx_locks.setdefault(key, threading.Lock())
+    if not lock.acquire(timeout=_XLSX_LOCK_TIMEOUT):
+        raise ServiceError(
+            "Workbook is being modified by another operation, retry shortly",
+            code="conflict",
+            status=409,
+            details={"path": key, "timeout_seconds": _XLSX_LOCK_TIMEOUT},
+        )
+    try:
+        yield
+    finally:
+        lock.release()
+
+
+def _coerce_xlsx_value(value: Any) -> Any:
+    """Turn the string sent by the cell editor back into a scalar (#153 A10).
+
+    The coercion is symmetric with :func:`backend.xlsx_reader._fmt`: a value
+    typed by the user comes back as a string, and Excel would have inferred a
+    type when typing the same thing. Recognised here:
+
+    * an empty cell -> ``None`` (clears it)
+    * ``1234`` / ``-1`` -> ``int``
+    * ``1.5`` / ``.5`` -> ``float``
+    * ``TRUE``/``FAUX`` (case-insensitive) -> ``bool``
+    * ``31/12/2026`` / ``31/12/2026 14:30`` -> ``date``/``datetime`` (FR)
+
+    Anything else stays text. A date-looking string typed with a leading
+    ``=`` is a formula and never reaches here as a date.
+    """
+    if not isinstance(value, str):
+        return value
+    text = value.strip()
+    if text == "":
+        return None
+    if _XLSX_INT_RE.match(text):
+        return int(text)
+    if _XLSX_FLOAT_RE.match(text):
+        return float(text)
+    lowered = text.lower()
+    if lowered in _XLSX_TRUE_LITERALS:
+        return True
+    if lowered in _XLSX_FALSE_LITERALS:
+        return False
+    if not _XLSX_FORMULA_RE.match(text):
+        parsed = _parse_fr_datetime(text)
+        if parsed is not None:
+            return parsed
+    return value
+
+
+def _parse_fr_datetime(text: str) -> date | datetime | None:
+    """Parse a FR-localised date/datetime, or return ``None``.
+
+    Accepts ``JJ/MM/AAAA`` and ``JJ/MM/AAAA HH:MM`` (also ``JJ-MM-AAAA``).
+    ``dayfirst`` is what makes ``01/02/2026`` the 1st of February rather than
+    the 2nd of January — the French convention.
+    """
+    if not _XLSX_DATE_RE.match(text):
+        return None
+    for fmt in ("%d/%m/%Y %H:%M", "%d/%m/%Y", "%d-%m-%Y %H:%M", "%d-%m-%Y"):
+        try:
+            return datetime.strptime(text, fmt)
+        except ValueError:
+            continue
+    return None
+
+
+def _write_cell(ws: Any, ref: str, value: Any, *, allow_formula: bool) -> None:
+    """Assign one cell, forcing text when it looks like a formula.
+
+    ``cell.data_type = "s"`` is what stops openpyxl from emitting ``<f>``: the
+    text is then stored as an inline/shared string and Excel shows it verbatim.
+    """
+    cell = ws[ref]
+    coerced = _coerce_xlsx_value(value)
+    cell.value = coerced
+    if not allow_formula and isinstance(coerced, str) and _XLSX_FORMULA_RE.match(coerced):
+        cell.data_type = "s"
+
+
+def edit_xlsx_cells(
+    vault_name: str,
+    path: str,
+    sheet: str,
+    cells: dict[str, Any],
+    *,
+    backup: bool = True,
+    allow_formula: bool = False,
+    force: bool = False,
+) -> dict[str, Any]:
+    """Apply a batch of cell edits to an ``.xlsx`` workbook.
+
+    Args:
+        vault_name: Name of the vault the workbook belongs to.
+        path: Vault-relative path of the ``.xlsx`` file.
+        sheet: Worksheet title to edit.
+        cells: Mapping of A1 references to new scalar values.
+        backup: Create a timestamped ``.bak`` before rewriting the archive.
+        allow_formula: Keep values starting with ``=``/``@`` as real formulas.
+            Off by default (#153 A4): a typed ``=cmd|…`` is a DDE payload when
+            the file is later opened in Excel.
+        force: Write even when the workbook carries features openpyxl drops
+            (slicers, form controls, connections, custom XML, signature, cached
+            formula results — see :data:`backend.xlsx_reader.LOSSY_PARTS`).
+
+    Raises:
+        ServiceError: ``not_found`` (404), ``read_only`` (403), ``conflict``
+        (409, concurrent write), ``xlsx_lossy_content`` (409, a lossy write was
+        attempted without ``force``) or ``invalid`` (400) for a bad sheet, cell
+        reference or value.
+    """
+    root = get_vault_root(vault_name)
+    _ensure_writable(root)
+    file_path = resolve_safe_path(root, path)
+
+    if not file_path.exists() or not file_path.is_file():
+        raise ServiceError(
+            f"File not found: {path}",
+            code="not_found",
+            status=404,
+            details={"vault": vault_name, "path": path},
+        )
+    if file_path.suffix.lower() != ".xlsx":
+        raise ServiceError(
+            f"Not an .xlsx file: {path}", code="invalid", status=400
+        )
+    if not cells:
+        raise ServiceError("No cells to update", code="invalid", status=400)
+    for ref in cells:
+        if not isinstance(ref, str) or not _XLSX_CELL_RE.match(ref):
+            raise ServiceError(
+                f"Invalid cell reference: {ref!r}", code="invalid", status=400
+            )
+
+    if not force:
+        from backend.xlsx_reader import inspect_workbook
+
+        lossy = inspect_workbook(file_path)
+        if lossy:
+            raise ServiceError(
+                "Saving this workbook would drop features ObsiGate cannot "
+                "preserve; retry with force=true after confirmation",
+                code="xlsx_lossy_content",
+                status=409,
+                details={"path": path, "features": lossy},
+            )
+
+    with _xlsx_write_lock(str(file_path)):
+        from openpyxl import load_workbook
+
+        try:
+            wb = load_workbook(file_path)
+        except Exception as exc:
+            raise ServiceError(
+                f"Cannot open workbook: {exc}", code="invalid", status=400
+            ) from exc
+        if sheet not in wb.sheetnames:
+            raise ServiceError(
+                f"Unknown sheet: {sheet}",
+                code="invalid",
+                status=400,
+                details={"sheets": wb.sheetnames},
+            )
+
+        rel_path = _rel(root, file_path)
+        if backup:
+            create_backup(file_path, vault_name, rel_path)
+
+        ws = wb[sheet]
+        for ref, value in cells.items():
+            _write_cell(ws, ref, value, allow_formula=allow_formula)
+        # #153 A2 — write beside the target then swap: a crash mid-save leaves
+        # the original workbook intact instead of a truncated archive.
+        tmp_path = file_path.with_name(f"{file_path.name}.{os.getpid()}.tmp")
+        try:
+            wb.save(tmp_path)
+            os.replace(tmp_path, file_path)
+        except Exception:
+            tmp_path.unlink(missing_ok=True)
+            raise
+
+    logger.info(f"XLSX cells saved: {vault_name}/{rel_path} [{sheet}] +{len(cells)}")
+    return {
+        "success": True,
+        "vault": vault_name,
+        "path": rel_path,
+        "size": len(cells),
+    }
+
+
 def append_to_file(
    vault_name: str,
    path: str,
@@ -10,6 +10,7 @@ No authentication required for public share views.
 import json
 import logging
 import secrets
+import threading
 from datetime import datetime, timedelta, timezone
 from pathlib import Path

@@ -17,6 +18,10 @@ logger = logging.getLogger("obsigate.share")

 SHARES_FILE = Path("data/shares.json")

+# ROADMAP #85 T10a — verrou autour des read-modify-write (perte de mises à
+# jour en cas de créations/accès/révocations concurrents).
+_lock = threading.RLock()
+

 def _read() -> dict:
    if not SHARES_FILE.exists():
@@ -41,26 +46,27 @@ def create_share(
    expires_in_hours: int | None = None,
 ) -> dict:
    """Create a new share token for a document."""
-    data = _read()
-    token = secrets.token_hex(32)  # 64-char hex token
+    with _lock:
+        data = _read()
+        token = secrets.token_hex(32)  # 64-char hex token

-    expires_at = None
-    if expires_in_hours:
-        expires_at = (datetime.now(timezone.utc) + timedelta(hours=expires_in_hours)).isoformat()
+        expires_at = None
+        if expires_in_hours:
+            expires_at = (datetime.now(timezone.utc) + timedelta(hours=expires_in_hours)).isoformat()

-    share = {
-        "id": token,
-        "token": token,
-        "vault": vault,
-        "path": path,
-        "created_by": created_by,
-        "created_at": datetime.now(timezone.utc).isoformat(),
-        "expires_at": expires_at,
-        "access_count": 0,
-        "last_accessed": None,
-    }
-    data["shares"][token] = share
-    _write(data)
+        share = {
+            "id": token,
+            "token": token,
+            "vault": vault,
+            "path": path,
+            "created_by": created_by,
+            "created_at": datetime.now(timezone.utc).isoformat(),
+            "expires_at": expires_at,
+            "access_count": 0,
+            "last_accessed": None,
+        }
+        data["shares"][token] = share
+        _write(data)
    logger.info(f"Created share for {vault}/{path} by {created_by}")
    return share

@@ -80,22 +86,24 @@ def get_share_by_token(token: str) -> dict | None:

 def record_access(token: str):
    """Increment access counter for a share."""
-    data = _read()
-    share = data["shares"].get(token)
-    if share:
-        share["access_count"] = share.get("access_count", 0) + 1
-        share["last_accessed"] = datetime.now(timezone.utc).isoformat()
-        _write(data)
+    with _lock:
+        data = _read()
+        share = data["shares"].get(token)
+        if share:
+            share["access_count"] = share.get("access_count", 0) + 1
+            share["last_accessed"] = datetime.now(timezone.utc).isoformat()
+            _write(data)


 def revoke_share(share_id: str) -> bool:
    """Revoke (delete) a share by its token."""
-    data = _read()
-    if share_id in data["shares"]:
-        del data["shares"][share_id]
-        _write(data)
-        logger.info(f"Revoked share {share_id}")
-        return True
+    with _lock:
+        data = _read()
+        if share_id in data["shares"]:
+            del data["shares"][share_id]
+            _write(data)
+            logger.info(f"Revoked share {share_id}")
+            return True
    return False


@@ -112,12 +120,13 @@ def list_shares(vault_filter: str | None = None) -> list:

 def update_shares_after_rename(vault: str, old_path: str, new_path: str):
    """Update all shares when a file is renamed."""
-    data = _read()
-    updated = False
-    for sid, s in data["shares"].items():
-        if s.get("vault") == vault and s.get("path") == old_path:
-            s["path"] = new_path
-            updated = True
-            logger.info(f"Updated share {sid}: {vault}/{old_path} -> {new_path}")
-    if updated:
-        _write(data)
+    with _lock:
+        data = _read()
+        updated = False
+        for sid, s in data["shares"].items():
+            if s.get("vault") == vault and s.get("path") == old_path:
+                s["path"] = new_path
+                updated = True
+                logger.info(f"Updated share {sid}: {vault}/{old_path} -> {new_path}")
+        if updated:
+            _write(data)
@@ -0,0 +1,54 @@
+"""Server-Sent Events manager (ROADMAP #85, tranche 4).
+
+Singleton extrait de :mod:`backend.main` sans changement de comportement :
+les routers montés par ``main`` partagent la même instance (les clients SSE
+connectés sur ``/api/events`` reçoivent les broadcasts émis depuis
+n'importe quel router).
+"""
+
+from __future__ import annotations
+
+import asyncio
+import json as _json
+import logging
+
+logger = logging.getLogger("obsigate")
+
+
+class SSEManager:
+    """Manages SSE client connections and broadcasts events."""
+
+    def __init__(self):
+        self._clients: list[asyncio.Queue] = []
+
+    async def connect(self) -> asyncio.Queue:
+        """Register a new SSE client and return its message queue."""
+        queue: asyncio.Queue = asyncio.Queue()
+        self._clients.append(queue)
+        logger.debug(f"SSE client connected (total: {len(self._clients)})")
+        return queue
+
+    def disconnect(self, queue: asyncio.Queue):
+        """Remove a disconnected SSE client."""
+        if queue in self._clients:
+            self._clients.remove(queue)
+        logger.debug(f"SSE client disconnected (total: {len(self._clients)})")
+
+    async def broadcast(self, event_type: str, data: dict):
+        """Send an event to all connected SSE clients."""
+        message = _json.dumps(data, ensure_ascii=False)
+        dead: list[asyncio.Queue] = []
+        for q in self._clients:
+            try:
+                q.put_nowait({"event": event_type, "data": message})
+            except asyncio.QueueFull:
+                dead.append(q)
+        for q in dead:
+            self.disconnect(q)
+
+    @property
+    def client_count(self) -> int:
+        return len(self._clients)
+
+
+sse_manager = SSEManager()
@@ -13,6 +13,7 @@ from backend.tools import connected as _connected  # noqa: F401  (registers conn
 from backend.tools import crawler as _crawler  # noqa: F401  (registers the site crawler)
 from backend.tools import documents as _documents  # noqa: F401  (registers document tools)
 from backend.tools import service as _service  # noqa: F401  (registers tools)
+from backend.tools import spreadsheets as _spreadsheets  # noqa: F401  (registers existing-workbook tools #153 A6)
 from backend.tools import web as _web  # noqa: F401  (registers web tools)
 from backend.tools.context import (
    ToolConfirmationRequired,
@@ -18,8 +18,10 @@ import csv as csv_lib
 import io
 import logging
 import re
-from typing import Any
-from xml.sax import saxutils
+from typing import Any, cast
+
+# saxutils.escape uniquement (échappement de chaînes, aucun parsing XML).
+from xml.sax import saxutils  # nosec B406

 from backend.services.errors import ServiceError
 from backend.services.mutations import save_raw_file
@@ -171,7 +173,9 @@ def _render_markdown_pdf(content: str, title: str) -> bytes | None:
            escape=False,
            plugins=["table", "strikethrough", "footnotes", "task_lists"],
        )
-        html = renderer(content)
+        # mistune 3.3 types `Markdown.__call__` as `str | list[...]` (le
+        # renderer HTML renvoie toujours `str` à l'exécution).
+        html = cast(str, renderer(content))
        return generate_pdf(build_pdf_html(html, title), title)
    except Exception as e:
        # WeasyPrint loads GTK lazily: a missing native library can surface at
@@ -52,6 +52,10 @@ _STEP_LABELS: dict[str, tuple[str, str | None]] = {
    "git_search_issues": ("git_issues", "query"),
    "git_get_file": ("git_file", "path"),
    "create_xlsx": ("xlsx_create", "path"),
+    "list_xlsx_sheets": ("xlsx_sheets", "path"),
+    "xlsx_to_markdown": ("xlsx_read", "path"),
+    "update_xlsx_cells": ("xlsx_update", "path"),
+    "append_xlsx_rows": ("xlsx_append", "path"),
    "create_docx": ("docx_create", "path"),
    "create_csv": ("csv_create", "path"),
    "create_pdf": ("pdf_create", "path"),
@@ -315,6 +315,61 @@ class DocxInput(BaseModel):
    overwrite: bool = Field(True, description="Replace an existing file (with backup)")


+class ListXlsxSheetsInput(BaseModel):
+    """List the sheets of an existing .xlsx workbook (#153 A6)."""
+
+    vault: str = Field(..., description="Vault name")
+    path: str = Field(..., description="Vault-relative path of the .xlsx file")
+
+
+class XlsxToMarkdownInput(BaseModel):
+    """Read one sheet of an existing .xlsx workbook as markdown (#153 A6)."""
+
+    vault: str = Field(..., description="Vault name")
+    path: str = Field(..., description="Vault-relative path of the .xlsx file")
+    sheet: str = Field(
+        "", description="Sheet name (empty = the first/active sheet)"
+    )
+
+
+class UpdateXlsxCellsInput(BaseModel):
+    """Batch-edit cells of an existing .xlsx workbook (#153 A6)."""
+
+    vault: str = Field(..., description="Vault name")
+    path: str = Field(..., description="Vault-relative path of the .xlsx file")
+    sheet: str = Field(..., description="Worksheet title to edit")
+    cells: dict[str, str | int | float | bool | None] = Field(
+        ..., description="A1 reference -> new value (max 500 per call)"
+    )
+    allow_formula: bool = Field(
+        False,
+        description="Store '='/'@' values as real formulas (off by default, DDE guard)",
+    )
+    force: bool = Field(
+        False,
+        description="Write even when features openpyxl cannot rewrite would be dropped",
+    )
+
+
+class AppendXlsxRowsInput(BaseModel):
+    """Append rows at the end of a sheet of an existing .xlsx (#153 A6)."""
+
+    vault: str = Field(..., description="Vault name")
+    path: str = Field(..., description="Vault-relative path of the .xlsx file")
+    sheet: str = Field(..., description="Worksheet title to extend")
+    rows: list[list[str | int | float | bool | None]] = Field(
+        ..., description="Rows of cell values, appended below the last used row (max 500)"
+    )
+    allow_formula: bool = Field(
+        False,
+        description="Store '='/'@' values as real formulas (off by default, DDE guard)",
+    )
+    force: bool = Field(
+        False,
+        description="Write even when features openpyxl cannot rewrite would be dropped",
+    )
+
+
 class CsvInput(BaseModel):
    """Create a .csv file in a vault from rows of cells."""

@@ -17,6 +17,7 @@ from __future__ import annotations
 import json
 import logging
 import os
+import threading
 from pathlib import Path

 logger = logging.getLogger("obsigate.tools.secrets")
@@ -34,6 +35,9 @@ TOOL_KEY_NAMES: tuple[str, ...] = (

 _SECRET_MARKERS = ("API_KEY", "TOKEN")

+# ROADMAP #85 T10a — verrou autour des read-modify-write du store de clés.
+_lock = threading.RLock()
+

 def _keys_file() -> Path:
    base = os.environ.get("OBSIGATE_DATA_DIR", "data")
@@ -89,21 +93,23 @@ def set_tool_key(name: str, value: str) -> None:
    if name not in TOOL_KEY_NAMES:
        raise ValueError(f"Clé non prise en charge: {name}")
    value = (value or "").strip()
-    keys = _read_keys()
-    if value:
-        keys[name] = value
-    else:
-        keys.pop(name, None)
-    _write_keys(keys)
+    with _lock:
+        keys = _read_keys()
+        if value:
+            keys[name] = value
+        else:
+            keys.pop(name, None)
+        _write_keys(keys)


 def delete_tool_key(name: str) -> bool:
    """Remove one key from the store; return True when it existed."""
    if name not in TOOL_KEY_NAMES:
        raise ValueError(f"Clé non prise en charge: {name}")
-    keys = _read_keys()
-    if name in keys:
-        del keys[name]
-        _write_keys(keys)
-        return True
+    with _lock:
+        keys = _read_keys()
+        if name in keys:
+            del keys[name]
+            _write_keys(keys)
+            return True
    return False
@@ -0,0 +1,286 @@
+"""Spreadsheet tools (#153 A6) — read and mutate existing ``.xlsx`` workbooks.
+
+Complements :mod:`backend.tools.documents` (``create_xlsx`` creates a *new*
+file; here the assistant can read and edit one that already exists):
+
+* ``list_xlsx_sheets``  — READ, sheet names + dimensions;
+* ``xlsx_to_markdown``  — READ, bounded markdown table for the LLM context;
+* ``update_xlsx_cells`` — WRITE, batch cell edits (wraps the guarded service);
+* ``append_xlsx_rows``  — WRITE, append whole rows at the end of a sheet.
+
+Mutation tools go through :func:`backend.services.mutations.edit_xlsx_cells`,
+which already carries the #153 P0 guards: per-file lock, atomic replace,
+formula neutralisation (``allow_formula`` opt-in) and the lossy-write 409.
+"""
+
+from __future__ import annotations
+
+import logging
+from pathlib import Path
+from typing import Any
+
+from backend.services.errors import ServiceError
+from backend.services.paths import resolve_safe_path
+from backend.services.vaults import get_vault_root
+from backend.tools.context import ToolContext, ToolError, ToolRisk
+from backend.tools.registry import tool
+from backend.tools.schemas import (
+    AppendXlsxRowsInput,
+    ListXlsxSheetsInput,
+    UpdateXlsxCellsInput,
+    XlsxToMarkdownInput,
+)
+
+logger = logging.getLogger("obsigate.tools.spreadsheets")
+
+# xlsx_to_markdown ceiling: a workbook is a data dump, not prose. The table is
+# for the LLM context, so both axes are bounded (same spirit as A5's index cap).
+MAX_MD_ROWS = 100
+MAX_MD_COLS = 20
+MAX_MD_CHARS = 20_000
+
+
+def _workbook_path(vault: str, path: str) -> Path:
+    """Resolve and validate a vault-relative ``.xlsx`` path."""
+    path = (path or "").strip()
+    if not path.lower().endswith(".xlsx"):
+        raise ToolError("Extension attendue : .xlsx", code="invalid_arguments")
+    try:
+        root = get_vault_root(vault)
+    except ServiceError as e:
+        raise ToolError(e.message, code=e.code, details=e.details) from e
+    return resolve_safe_path(root, path)
+
+
+def _map_service_error(e: ServiceError) -> ToolError:
+    return ToolError(e.message, code=e.code, details=e.details)
+
+
+@tool(
+    name="list_xlsx_sheets",
+    description=(
+        "List the sheets of an .xlsx workbook with their dimensions "
+        "(rows x columns) and whether the display caps truncate them. "
+        "Use before editing to pick the right sheet name."
+    ),
+    input_model=ListXlsxSheetsInput,
+    risk=ToolRisk.READ,
+    requires_vault=True,
+)
+def list_xlsx_sheets(ctx: ToolContext, params: ListXlsxSheetsInput) -> dict[str, Any]:
+    """Return sheet names and extents of the workbook."""
+    from backend.xlsx_reader import MAX_COLS, MAX_ROWS, _sheet_extent
+
+    file_path = _workbook_path(params.vault, params.path)
+    try:
+        from openpyxl import load_workbook
+
+        wb = load_workbook(str(file_path), read_only=True, data_only=True)
+    except ServiceError as e:
+        raise _map_service_error(e) from e
+    except Exception as e:
+        raise ToolError(f"Classeur illisible: {e}", code="invalid") from e
+    try:
+        sheets = []
+        for ws in wb.worksheets:
+            total_rows, total_cols = _sheet_extent(ws)
+            sheets.append(
+                {
+                    "name": ws.title,
+                    "total_rows": total_rows,
+                    "total_cols": total_cols,
+                    "truncated": total_rows > MAX_ROWS or total_cols > MAX_COLS,
+                }
+            )
+        return {"vault": params.vault, "path": params.path, "sheets": sheets}
+    finally:
+        wb.close()
+
+
+@tool(
+    name="xlsx_to_markdown",
+    description=(
+        "Read a sheet of an .xlsx workbook as a bounded markdown table "
+        "(up to 100 rows x 20 columns). Use to inspect spreadsheet data "
+        "before answering or editing."
+    ),
+    input_model=XlsxToMarkdownInput,
+    risk=ToolRisk.READ,
+    requires_vault=True,
+)
+def xlsx_to_markdown(ctx: ToolContext, params: XlsxToMarkdownInput) -> dict[str, Any]:
+    """Render one sheet as a markdown table for the LLM context."""
+    from openpyxl import load_workbook
+
+    from backend.xlsx_reader import _fmt
+
+    file_path = _workbook_path(params.vault, params.path)
+    try:
+        wb = load_workbook(str(file_path), read_only=True, data_only=True)
+    except ServiceError as e:
+        raise _map_service_error(e) from e
+    except Exception as e:
+        raise ToolError(f"Classeur illisible: {e}", code="invalid") from e
+    try:
+        if params.sheet:
+            if params.sheet not in wb.sheetnames:
+                raise ToolError(
+                    f"Feuille introuvable: {params.sheet}", code="not_found"
+                )
+            ws = wb[params.sheet]
+        else:
+            ws = wb.active
+        title = ws.title
+        rows: list[list[str]] = []
+        truncated = False
+        for row in ws.iter_rows(
+            min_row=1, max_row=MAX_MD_ROWS, max_col=MAX_MD_COLS, values_only=True
+        ):
+            cells = [_fmt(v) for v in row]
+            if not any(c.strip() for c in cells):
+                continue
+            rows.append(cells)
+        # Real tail beyond the caps? Probe one row further.
+        probe = list(
+            ws.iter_rows(
+                min_row=MAX_MD_ROWS + 1,
+                max_row=MAX_MD_ROWS + 1,
+                max_col=MAX_MD_COLS,
+                values_only=True,
+            )
+        )
+        if any(any(str(v or "").strip() for v in r) for r in probe):
+            truncated = True
+    finally:
+        wb.close()
+
+    lines: list[str] = []
+    if rows:
+        header = rows[0]
+        lines.append("| " + " | ".join(header) + " |")
+        lines.append("|" + "|".join("---" for _ in header) + "|")
+        for row in rows[1:]:
+            lines.append("| " + " | ".join(row) + " |")
+    table = "\n".join(lines)[:MAX_MD_CHARS]
+
+    return {
+        "vault": params.vault,
+        "path": params.path,
+        "sheet": title,
+        "rows": len(rows),
+        "cols": max((len(r) for r in rows), default=0),
+        "truncated": truncated,
+        "markdown": table,
+    }
+
+
+@tool(
+    name="update_xlsx_cells",
+    description=(
+        "Edit cells of an existing .xlsx workbook. ``cells`` maps A1 "
+        "references to new values (max 500). A value starting with '=' or "
+        "'@' is stored as TEXT unless allow_formula is set (DDE guard). "
+        "Editing a workbook carrying features openpyxl cannot rewrite "
+        "requires force=true (cached formula results, slicers…)."
+    ),
+    input_model=UpdateXlsxCellsInput,
+    risk=ToolRisk.WRITE,
+    requires_vault=True,
+)
+def update_xlsx_cells(ctx: ToolContext, params: UpdateXlsxCellsInput) -> dict[str, Any]:
+    """Wrap the guarded cell-edit service."""
+    from backend.services.mutations import edit_xlsx_cells
+
+    if not params.cells:
+        raise ToolError("Aucune cellule fournie", code="invalid_arguments")
+    try:
+        result = edit_xlsx_cells(
+            params.vault,
+            params.path,
+            params.sheet,
+            dict(params.cells),
+            allow_formula=params.allow_formula,
+            force=params.force,
+        )
+    except ServiceError as e:
+        raise _map_service_error(e) from e
+    return {
+        "status": "ok",
+        "vault": result["vault"],
+        "path": result["path"],
+        "sheet": params.sheet,
+        "cells": len(params.cells),
+    }
+
+
+@tool(
+    name="append_xlsx_rows",
+    description=(
+        "Append rows at the end of a sheet of an existing .xlsx workbook. "
+        "Values are typed like in the viewer (numbers, TRUE/FALSE, FR dates "
+        "JJ/MM/AAAA). The workbook is rewritten atomically with a backup."
+    ),
+    input_model=AppendXlsxRowsInput,
+    risk=ToolRisk.WRITE,
+    requires_vault=True,
+)
+def append_xlsx_rows(ctx: ToolContext, params: AppendXlsxRowsInput) -> dict[str, Any]:
+    """Append whole rows below the last used row of the sheet."""
+    from openpyxl import load_workbook
+    from openpyxl.utils import get_column_letter
+
+    from backend.services.mutations import _coerce_xlsx_value, edit_xlsx_cells
+
+    if not params.rows:
+        raise ToolError("Aucune ligne fournie", code="invalid_arguments")
+    if len(params.rows) > 500:
+        raise ToolError("Trop de lignes (max 500)", code="invalid_arguments")
+
+    file_path = _workbook_path(params.vault, params.path)
+    try:
+        wb = load_workbook(str(file_path), read_only=True, data_only=True)
+        try:
+            if params.sheet not in wb.sheetnames:
+                raise ToolError(
+                    f"Feuille introuvable: {params.sheet}", code="not_found"
+                )
+            ws = wb[params.sheet]
+            first_free = (ws.max_row or 0) + 1
+        finally:
+            wb.close()
+    except ServiceError as e:
+        raise _map_service_error(e) from e
+    except ToolError:
+        raise
+    except Exception as e:
+        raise ToolError(f"Classeur illisible: {e}", code="invalid") from e
+
+    cells: dict[str, Any] = {}
+    for i, row in enumerate(params.rows):
+        for j, value in enumerate(row):
+            if value is None or (isinstance(value, str) and not value.strip()):
+                continue
+            ref = f"{get_column_letter(j + 1)}{first_free + i}"
+            cells[ref] = _coerce_xlsx_value(value)
+    if not cells:
+        raise ToolError("Aucune valeur fournie", code="invalid_arguments")
+
+    try:
+        result = edit_xlsx_cells(
+            params.vault,
+            params.path,
+            params.sheet,
+            cells,
+            allow_formula=params.allow_formula,
+            force=params.force,
+        )
+    except ServiceError as e:
+        raise _map_service_error(e) from e
+    return {
+        "status": "ok",
+        "vault": result["vault"],
+        "path": result["path"],
+        "sheet": params.sheet,
+        "rows": len(params.rows),
+        "first_row": first_free,
+    }
@@ -22,7 +22,7 @@ Exemples :
 from __future__ import annotations

 import os
-import subprocess
+import subprocess  # nosec B404
 from pathlib import Path

 _ROOT = Path(__file__).resolve().parent.parent          # racine du dépôt ObsiGate
@@ -34,7 +34,8 @@ _ENV_VAR = "OBSIGATE_VERSION"
 def _run_git(args: list[str]) -> str:
    """Run a git command in the repo root; return stdout (stripped) or ''."""
    try:
-        result = subprocess.run(
+        # argv fixe (git + args internes), sans shell : pas d'injection.
+        result = subprocess.run(  # nosec B404 B603 B607
            ["git", *args],
            cwd=str(_ROOT),
            capture_output=True,
@@ -280,7 +280,8 @@ class VaultWatcher:
        for observer in self.observers.values():
            try:
                observer.join(timeout=5)
-            except Exception:  # nosec B110 — best-effort shutdown, ignore failures
+            # best-effort shutdown, ignore failures (B110) :
+            except Exception:  # nosec B110
                pass
        self.observers.clear()
        logger.info("VaultWatcher stopped")
@@ -0,0 +1,28 @@
+"""Shared VaultWatcher handle (ROADMAP #85, tranche 8).
+
+Holder extrait de :mod:`backend.main` sans changement de comportement : le
+lifespan de ``main`` y dépose l'instance (``set_watcher``) et l'y reprend à
+l'extinction ; le router ``vaults`` la consulte via :func:`get_watcher`
+(démarrage/arrêt de surveillance à l'ajout/retrait dynamique de vault,
+état dans ``/api/vaults/status``).
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from backend.watcher import VaultWatcher
+
+_watcher: VaultWatcher | None = None
+
+
+def get_watcher() -> VaultWatcher | None:
+    """Return the shared VaultWatcher instance (``None`` if disabled)."""
+    return _watcher
+
+
+def set_watcher(watcher: VaultWatcher | None) -> None:
+    """Store (or clear) the shared VaultWatcher instance."""
+    global _watcher
+    _watcher = watcher
@@ -26,6 +26,7 @@ import json
 import logging
 import os
 import socket
+import threading
 import uuid
 from datetime import datetime, timezone
 from pathlib import Path
@@ -144,6 +145,12 @@ def _read_secrets() -> dict:
        return {}


+# ROADMAP #85 T10a — verrou autour des read-modify-write des deux stores
+# (webhooks + secrets) : perte de mises à jour en cas de mutations
+# concurrentes.
+_lock = threading.RLock()
+
+
 def _write_secrets(secrets: dict):
    WEBHOOK_SECRETS_FILE.parent.mkdir(parents=True, exist_ok=True)
    tmp = WEBHOOK_SECRETS_FILE.with_suffix(".tmp")
@@ -156,12 +163,13 @@ def _write_secrets(secrets: dict):


 def _store_secret(wh_id: str, secret: str | None) -> None:
-    secrets = _read_secrets()
-    if secret:
-        secrets[wh_id] = secret
-    else:
-        secrets.pop(wh_id, None)
-    _write_secrets(secrets)
+    with _lock:
+        secrets = _read_secrets()
+        if secret:
+            secrets[wh_id] = secret
+        else:
+            secrets.pop(wh_id, None)
+        _write_secrets(secrets)


 def _get_secret(wh: dict) -> str | None:
@@ -189,52 +197,55 @@ def get_webhooks() -> list:

 def create_webhook(name: str, url: str, events: list[str], secret: str | None = None) -> dict:
    validate_webhook_url(url)
-    webhooks = _read()
-    wh_id = str(uuid.uuid4())
-    wh = {
-        "id": wh_id,
-        "name": name,
-        "url": url,
-        "events": [e for e in events if e in VALID_EVENTS],
-        "enabled": True,
-        "created_at": datetime.now(timezone.utc).isoformat(),
-        "last_fired_at": None,
-    }
-    webhooks.append(wh)
-    _write(webhooks)
-    if secret:
-        _store_secret(wh_id, secret)
+    with _lock:
+        webhooks = _read()
+        wh_id = str(uuid.uuid4())
+        wh = {
+            "id": wh_id,
+            "name": name,
+            "url": url,
+            "events": [e for e in events if e in VALID_EVENTS],
+            "enabled": True,
+            "created_at": datetime.now(timezone.utc).isoformat(),
+            "last_fired_at": None,
+        }
+        webhooks.append(wh)
+        _write(webhooks)
+        if secret:
+            _store_secret(wh_id, secret)
    logger.info(f"Created webhook '{name}' → {url}")
    return _public_view(wh)


 def update_webhook(wh_id: str, updates: dict) -> dict | None:
-    webhooks = _read()
-    for wh in webhooks:
-        if wh["id"] == wh_id:
-            if updates.get("url"):
-                validate_webhook_url(updates["url"])
-            if "secret" in updates:
-                _store_secret(wh_id, updates["secret"])
-            safe_updates = {
-                k: v for k, v in updates.items()
-                if k not in ("id", "secret")
-            }
-            wh.update(safe_updates)
-            _write(webhooks)
-            return _public_view(wh)
+    with _lock:
+        webhooks = _read()
+        for wh in webhooks:
+            if wh["id"] == wh_id:
+                if updates.get("url"):
+                    validate_webhook_url(updates["url"])
+                if "secret" in updates:
+                    _store_secret(wh_id, updates["secret"])
+                safe_updates = {
+                    k: v for k, v in updates.items()
+                    if k not in ("id", "secret")
+                }
+                wh.update(safe_updates)
+                _write(webhooks)
+                return _public_view(wh)
    return None


 def delete_webhook(wh_id: str) -> bool:
-    webhooks = _read()
-    new_list = [wh for wh in webhooks if wh["id"] != wh_id]
-    if len(new_list) == len(webhooks):
-        return False
-    _write(new_list)
-    secrets = _read_secrets()
-    if secrets.pop(wh_id, None) is not None:
-        _write_secrets(secrets)
+    with _lock:
+        webhooks = _read()
+        new_list = [wh for wh in webhooks if wh["id"] != wh_id]
+        if len(new_list) == len(webhooks):
+            return False
+        _write(new_list)
+        secrets = _read_secrets()
+        if secrets.pop(wh_id, None) is not None:
+            _write_secrets(secrets)
    return True


@@ -0,0 +1,446 @@
+"""Render ``.xlsx`` workbooks as HTML tables for the viewer (#xlsx).
+
+Read-only: formulas are shown as their text (``data_only=False``) so a
+round-trip through the viewer never depends on Excel's cached values.
+Write-side lives in ``backend.services.mutations.edit_xlsx_cells``.
+
+:func:`inspect_workbook` lists the workbook features that an openpyxl
+round-trip would drop (#153 A1) so the UI can warn before saving.
+"""
+
+from __future__ import annotations
+
+import html
+import logging
+import re
+import zipfile
+from datetime import date, datetime
+from pathlib import Path
+from typing import Any
+
+from openpyxl import load_workbook
+from openpyxl.utils import get_column_letter
+
+logger = logging.getLogger("obsigate.xlsx_reader")
+
+# ponytail: hard caps bound the rendered grid (500 rows x 40 cols per sheet).
+# Raise them, or paginate per sheet, if a real workbook needs more.
+MAX_ROWS = 500
+MAX_COLS = 40
+
+# #153 A9 — window size served by ``read_sheet_window()`` (lazy per-sheet
+# loading). The endpoint is bounded so a single request can never ask for the
+# whole workbook back in one JSON payload; the UI pages through the rest.
+MAX_WINDOW_ROWS = 1_000
+DEFAULT_WINDOW_ROWS = 200
+
+# #153 A1 — workbook parts openpyxl does not re-serialize on load+save.
+# Verified against openpyxl 3.1.5: charts, images, drawings and pivot tables
+# DO survive the round-trip, so they are deliberately absent from this map.
+LOSSY_PARTS: dict[str, tuple[str, ...]] = {
+    "slicers": ("xl/slicers/", "xl/slicerCaches/", "xl/timelines/"),
+    "form_controls": ("xl/ctrlProps/", "xl/activeX/"),
+    "connections": ("xl/queryTables/", "xl/connections.xml"),
+    "custom_xml": ("customXml/",),
+    "signature": ("_xmlsignatures/",),
+    "rich_comments": ("xl/threadedComments/", "xl/persons/"),
+    "macros": ("xl/vbaProject.bin",),
+}
+
+# A formula cell carrying its last computed result: ``<f>…</f><v>…</v>``.
+# openpyxl writes an EMPTY ``<v></v>`` itself, hence the ``[^<]`` guard: only a
+# non-empty value counts. openpyxl keeps the formula but drops the cached result,
+# so any reader using ``data_only=True`` (pandas, converters) sees ``None`` until
+# Excel recalculates.
+_CACHED_FORMULA_RE = re.compile(rb"<f[ >][^<]*</f>\s*<v>[^<]")
+
+# Sheet XML scanned by the cached-formula probe (CPU guard, like MAX_REPLACE_FILE_BYTES).
+_MAX_PROBE_BYTES = 8_000_000
+
+# #153 A5 — ceiling on the text handed to the TF-IDF / semantic index. A workbook
+# is a data dump, not prose: indexing every cell would flood the inverted index
+# and bury the notes. Sheet names + the first rows are enough to make a
+# spreadsheet findable by its headers.
+MAX_INDEX_CHARS = 5_000
+_INDEX_ROWS_PER_SHEET = 20
+MAX_INDEX_SHEETS = 20
+
+
+def _fmt(value: Any) -> str:
+    if value is None:
+        return ""
+    if isinstance(value, datetime):
+        return value.strftime("%Y-%m-%d %H:%M")
+    if isinstance(value, date):
+        return value.isoformat()
+    return str(value)
+
+
+def _trim(grid: list[list[str]]) -> list[list[str]]:
+    """Drop trailing empty rows and columns (openpyxl pads to max_col)."""
+    while grid and not any(grid[-1]):
+        grid.pop()
+    if not grid:
+        return grid
+    width = 0
+    for row in grid:
+        for i in range(len(row) - 1, -1, -1):
+            if row[i]:
+                width = max(width, i + 1)
+                break
+    return [row[:width] for row in grid]
+
+
+def _cell_cached(cached: list[list[str]] | None, r: int, c: int) -> str:
+    """Return the cached result for a 0-based cell, or ``""``.
+
+    The shadow grid is read positionally and may be narrower than the formula
+    grid (``_trim`` collapses the trailing empty columns of each grid
+    independently), so every lookup is bounds-checked rather than assumed.
+    """
+    if not cached or r >= len(cached):
+        return ""
+    row = cached[r]
+    return row[c] if c < len(row) else ""
+
+
+def _table(
+    grid: list[list[str]],
+    cached: list[list[str]] | None = None,
+    row_offset: int = 0,
+) -> str:
+    """Render a grid as an HTML table.
+
+    ``cached`` is the same grid read with ``data_only=True`` (#153 A12): where a
+    formula cell still carries its last computed result, it is shown as a
+    discreet second line (``<span class="xlsx-cached">``) so the user sees the
+    number Excel last calculated instead of only the formula text. The span
+    carries ``data-cached-value`` and is titled client-side from
+    ``xlsx.cached_value_title`` — the backend never emits UI text.
+
+    ``row_offset`` is the number of rows skipped before this grid (#153 A9): the
+    row numbers and the ``data-cell`` references must stay the real A1
+    coordinates of the sheet, not of the window.
+    """
+    if not grid:
+        return "<p><em>Feuille vide</em></p>"
+    n_cols = max(len(row) for row in grid)
+    out = [
+        (
+            '<div class="csv-table-wrapper"><table class="csv-table xlsx-table">'
+            '<thead><tr><th class="xlsx-corner"></th>'
+        )
+    ]
+    out += [f"<th>{get_column_letter(c)}</th>" for c in range(1, n_cols + 1)]
+    out.append("</tr></thead><tbody>")
+    for r, row in enumerate(grid, start=row_offset + 1):
+        out.append(f'<tr><th class="xlsx-rownum">{r}</th>')
+        for c, val in enumerate(row, start=1):
+            ref = f"{get_column_letter(c)}{r}"
+            # The cached result only makes sense for a formula cell: on a plain
+            # value cell the two reads are identical and showing both would
+            # duplicate the text.
+            shadow = ""
+            if cached is not None and val.startswith("="):
+                # `c` is 1-based (A1 notation) and `r` too, while the grid is
+                # 0-based: translate both.
+                cval = _cell_cached(cached, r - 1, c - 1)
+                if cval and cval != val:
+                    # The tooltip is translated client-side from
+                    # `xlsx.cached_value_title`; never hardcode UI text here.
+                    shadow = (
+                        f'<span class="xlsx-cached" data-cached-value="1">'
+                        f"{html.escape(cval)}</span>"
+                    )
+            out.append(
+                f'<td data-cell="{ref}">{html.escape(val)}{shadow}</td>'
+            )
+        out.append("</tr>")
+    out.append("</tbody></table></div>")
+    return "".join(out)
+
+
+def _has_cached_formulas(zf: zipfile.ZipFile) -> bool:
+    """True when at least one formula cell still carries its computed value."""
+    budget = _MAX_PROBE_BYTES
+    for name in zf.namelist():
+        if not name.startswith("xl/worksheets/sheet") or not name.endswith(".xml"):
+            continue
+        try:
+            with zf.open(name) as fh:
+                while budget > 0:
+                    chunk = fh.read(65536)
+                    if not chunk:
+                        break
+                    budget -= len(chunk)
+                    if _CACHED_FORMULA_RE.search(chunk):
+                        return True
+        except (KeyError, OSError, zipfile.BadZipFile):
+            continue
+    return False
+
+
+def inspect_workbook(file_path: Path) -> list[str]:
+    """Return the sorted keys of :data:`LOSSY_PARTS` present in *file_path*.
+
+    Read-only inspection of the OPC package (central directory + a bounded scan
+    of the sheet XML). Never raises: an unreadable or encrypted workbook simply
+    yields ``[]`` and the save path keeps its current behaviour.
+
+    ``cached_values`` is a synthetic key: openpyxl keeps the formula but drops
+    the cached result, so the workbook stays correct once Excel recalculates it.
+    """
+    try:
+        with zipfile.ZipFile(file_path) as zf:
+            names = set(zf.namelist())
+            found = {
+                key
+                for key, prefixes in LOSSY_PARTS.items()
+                if any(name.startswith(prefix) for name in names for prefix in prefixes)
+            }
+            if _has_cached_formulas(zf):
+                found.add("cached_values")
+            return sorted(found)
+    except (OSError, zipfile.BadZipFile):
+        return []
+
+
+def render_sheets(file_path: Path) -> list[dict[str, Any]]:
+    """Return one dict per sheet: ``{name, html, rows, cols, total_*, truncated}``.
+
+    Reads the workbook twice: once with ``data_only=False`` for the formulas
+    (what the user must edit) and, when any formula carries a cached result
+    (#153 A12), once with ``data_only=True`` to show what Excel last computed.
+    The second pass is skipped entirely when the archive holds no cached value,
+    so the common case still costs a single load.
+
+    ``total_rows``/``total_cols`` are the dimensions the sheet declares and
+    ``truncated`` says whether the hard caps actually cut it (#153 A8) — the
+    viewer needs both to stop silently hiding the tail of a sheet.
+    """
+    wb = load_workbook(str(file_path), read_only=True, data_only=False)
+    try:
+        formulas = [_sheet_grid(ws) for ws in wb.worksheets]
+        titles = [ws.title for ws in wb.worksheets]
+        extents = [_sheet_extent(ws) for ws in wb.worksheets]
+    finally:
+        wb.close()
+
+    cached: list[list[list[str]]] | None = None
+    if _has_cached_values(file_path):
+        cached = _read_cached_grids(file_path, titles)
+
+    sheets = []
+    for i, title in enumerate(titles):
+        grid = _trim(formulas[i])
+        # The shadow grid is NOT trimmed independently: _trim drops the
+        # trailing empty columns of each grid on its own width, which would
+        # shift every cached value left of its formula. Indexing it
+        # positionally against the untrimmed grid keeps the two aligned.
+        shadow = cached[i] if cached is not None and i < len(cached) else None
+        total_rows, total_cols = extents[i]
+        sheets.append(
+            {
+                "name": title,
+                "html": _table(grid, shadow),
+                "rows": len(grid),
+                "cols": max((len(r) for r in grid), default=0),
+                "total_rows": total_rows,
+                "total_cols": total_cols,
+                # Coverage, not display size: `rows`/`cols` are post-trim (a
+                # sheet of 3 filled cells in a 500-row block renders 1x1), and
+                # the client must announce the cap it stopped at, not how many
+                # cells happen to be non-empty.
+                "max_rows": MAX_ROWS,
+                "max_cols": MAX_COLS,
+                # A sheet is truncated when the caps, not the trailing blanks,
+                # decided its shape: comparing against the *rendered* size would
+                # flag every sheet carrying a few empty formatted rows.
+                "truncated": total_rows > MAX_ROWS or total_cols > MAX_COLS,
+            }
+        )
+    return sheets
+
+
+def read_sheet_window(
+    file_path: Path,
+    sheet: str,
+    offset: int = 0,
+    limit: int = DEFAULT_WINDOW_ROWS,
+) -> dict[str, Any] | None:
+    """Return a window of rows of one sheet, or ``None`` if the sheet is unknown.
+
+    Backs the lazy per-sheet loading of #153 A9: the viewer asks for the rows
+    it is about to display instead of shipping every sheet in the initial file
+    payload. ``offset`` is 0-based; the row numbers and the ``data-cell``
+    references in the returned ``html`` are the real A1 coordinates of the
+    sheet, so a window is indistinguishable from a full render.
+
+    ``limit`` is clamped to :data:`MAX_WINDOW_ROWS`. Raises nothing: an unknown
+    sheet yields ``None`` and a broken workbook propagates the caller's usual
+    500.
+    """
+    offset = max(int(offset), 0)
+    limit = min(max(int(limit), 1), MAX_WINDOW_ROWS)
+
+    wb = load_workbook(str(file_path), read_only=True, data_only=False)
+    try:
+        if sheet not in wb.sheetnames:
+            return None
+        ws = wb[sheet]
+        total_rows, total_cols = _sheet_extent(ws)
+        grid = _trim(
+            _sheet_grid(ws, min_row=offset + 1, max_row=offset + limit)
+        )
+    finally:
+        wb.close()
+
+    shadow: list[list[str]] | None = None
+    # Same A12 rule as the full render: the second read only happens when the
+    # archive really holds cached results.
+    if _has_cached_values(file_path):
+        shadow = _read_cached_window(file_path, sheet, offset, limit)
+    return {
+        "sheet": sheet,
+        "offset": offset,
+        "limit": limit,
+        "rows": len(grid),
+        "cols": max((len(r) for r in grid), default=0),
+        "total_rows": total_rows,
+        "total_cols": total_cols,
+        "max_rows": MAX_ROWS,
+        "max_cols": MAX_COLS,
+        "truncated": total_rows > MAX_ROWS or total_cols > MAX_COLS,
+        "has_more": offset + len(grid) < total_rows,
+        "html": _table(grid, shadow, row_offset=offset),
+    }
+
+
+def _read_cached_window(
+    file_path: Path, sheet: str, offset: int, limit: int
+) -> list[list[str]] | None:
+    """``data_only=True`` grid for one window, or ``None`` if unavailable.
+
+    Best effort like :func:`_read_cached_grids`: a workbook Excel opens but
+    openpyxl cannot re-read must still display (formulas only).
+    """
+    try:
+        wb = load_workbook(str(file_path), read_only=True, data_only=True)
+    except Exception:
+        return None
+    try:
+        if sheet not in wb.sheetnames:
+            return None
+        return _sheet_grid(
+            wb[sheet], min_row=offset + 1, max_row=offset + limit
+        )
+    except Exception:
+        logger.debug("xlsx cached window unavailable", exc_info=True)
+        return None
+    finally:
+        wb.close()
+
+
+def _sheet_extent(ws: Any) -> tuple[int, int]:
+    """Rows and columns the worksheet declares, never negative.
+
+    ``max_row``/``max_column`` come from the sheet's dimension record; a
+    hand-edited file may omit it, hence the defensive coercion.
+    """
+    try:
+        rows = max(int(getattr(ws, "max_row", 0) or 0), 0)
+    except (TypeError, ValueError):
+        rows = 0
+    try:
+        cols = max(int(getattr(ws, "max_column", 0) or 0), 0)
+    except (TypeError, ValueError):
+        cols = 0
+    return rows, cols
+
+
+def _sheet_grid(
+    ws: Any, min_row: int = 1, max_row: int = MAX_ROWS, max_col: int = MAX_COLS
+) -> list[list[str]]:
+    """Read a worksheet window into a grid of formatted strings, bounded by the caps."""
+    return [
+        [_fmt(v) for v in row]
+        for row in ws.iter_rows(
+            min_row=min_row, max_row=max_row, max_col=max_col, values_only=True
+        )
+    ]
+
+
+def _has_cached_values(file_path: Path) -> bool:
+    """True when the archive holds at least one ``<f>…</f><v>…</v>``."""
+    try:
+        with zipfile.ZipFile(file_path) as zf:
+            return _has_cached_formulas(zf)
+    except (OSError, zipfile.BadZipFile):
+        return False
+
+
+def _read_cached_grids(
+    file_path: Path, titles: list[str]
+) -> list[list[list[str]]] | None:
+    """Read every sheet with ``data_only=True`` (what Excel last computed).
+
+    Best effort: returns ``None`` on any failure so the viewer falls back to the
+    formula-only rendering. A workbook Excel opens but openpyxl cannot re-read
+    must still display.
+    """
+    try:
+        wb = load_workbook(str(file_path), read_only=True, data_only=True)
+    except Exception:
+        return None
+    try:
+        grids = [_sheet_grid(ws) for ws in wb.worksheets]
+        if [ws.title for ws in wb.worksheets] != titles:
+            return None
+        return grids
+    except Exception:
+        logger.debug("xlsx cached values unavailable", exc_info=True)
+        return None
+    finally:
+        wb.close()
+
+
+def extract_indexable_text(file_path: Path) -> str:
+    """Return searchable text for the TF-IDF / semantic index (#153 A5).
+
+    Sheet names plus the first :data:`_INDEX_ROWS_PER_SHEET` rows of each
+    sheet, capped at :data:`MAX_INDEX_CHARS`. Rows are tab-joined so a search
+    for a header matches the sheet it belongs to.
+
+    Never raises: a corrupt, encrypted or unsupported workbook yields ``""`` so
+    the file still gets indexed by name (same contract as :func:`inspect_workbook`).
+    """
+    chunks: list[str] = []
+    budget = MAX_INDEX_CHARS
+    try:
+        wb = load_workbook(str(file_path), read_only=True, data_only=True)
+    except Exception:
+        # Encrypted (BadZipFile) or not a real workbook: name-only indexing.
+        return ""
+    try:
+        for ws in wb.worksheets[:MAX_INDEX_SHEETS]:
+            if budget <= 0:
+                break
+            # The sheet title alone is a strong signal ("Recettes", "Budget").
+            block = [ws.title]
+            for row in ws.iter_rows(
+                min_row=1, max_row=_INDEX_ROWS_PER_SHEET, max_col=MAX_COLS, values_only=True
+            ):
+                cells = [_fmt(v) for v in row]
+                # Skip blank rows instead of emitting runs of tabs.
+                if not any(c.strip() for c in cells):
+                    continue
+                block.append("\t".join(cells).rstrip())
+            text = "\n".join(block)
+            chunks.append(text[:budget])
+            budget -= len(text)
+    except Exception:
+        # Truncated but still useful: keep whatever was collected.
+        pass
+    finally:
+        wb.close()
+    return "\n".join(c for c in chunks if c).strip()
@@ -2626,7 +2626,7 @@ dependencies = [

 [[package]]
 name = "obsigate-desktop"
-version = "2.22.0"
+version = "2.33.0"
 dependencies = [
 "chrono",
 "env_logger",
@@ -1,6 +1,6 @@
 [package]
 name = "obsigate-desktop"
-version = "2.22.0"
+version = "2.33.0"
 description = "ObsiGate Desktop — Porte d'entrée native pour vos vaults Obsidian"
 authors = ["Bruno Charest"]
 edition = "2021"
@@ -1,7 +1,7 @@
 {
  "$schema": "https://raw.githubusercontent.com/nicedoc/obsigate/main/desktop/tauri.conf.schema.json",
  "productName": "ObsiGate",
-  "version": "2.22.0",
+  "version": "2.33.0",
  "identifier": "com.obsigate.desktop",
  "build": {
    "frontendDist": "../frontend",
@@ -53,7 +53,12 @@ services:
      - OBSIGATE_AUTH_ENABLED=true
      - OBSIGATE_ADMIN_USER=admin
      # OBSIGATE_ADMIN_PASSWORD → .env
-      # OBSIGATE_SECURE_COOKIES=true   # si derrière reverse proxy HTTPS
+      # OBSIGATE_SECURE_COOKIES : auto par défaut (Secure si https, sinon
+      # pas de flag) — forcer à true uniquement si le proxy termine TLS
+      # sans X-Forwarded-Proto (avec TRUST_PROXY, l'auto suffit).
+      # Reverse proxy devant l'app : IPs d'audit réelles (BUG-030) et
+      # X-Forwarded-Proto honoré pour les cookies Secure (auto).
+      - OBSIGATE_TRUST_PROXY=true
      - OLLAMA_BASE_URL=http://ollama:11434/v1
      - OLLAMA_MODEL=qwen2.5-coder:1.5b
    env_file:
@@ -6,7 +6,7 @@ vaults Obsidian et raccourcis essentiels.

 > **Public :** tous les utilisateurs · **Durée de lecture :** ~10 min
 > **Voir aussi :** [Déploiement Docker](./DEPLOIEMENT_DOCKER.md) ·
-> [Recherche, PDF & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md) ·
+> [Recherche, PDF, Excel & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md) ·
 > [API REST](./API_REST.md)

 ---
@@ -185,7 +185,7 @@ des **onglets** (avec possibilité de vue multi-panneaux / split view).
 La recherche est un point fort d'ObsiGate : index inversé TF-IDF, stemming
 français, normalisation des accents, facettes et pagination. La syntaxe complète
 (`tag:`, `#`, `vault:`, `title:`, `path:`, `ext:`, phrases exactes) est décrite
-dans le [Guide Recherche, PDF & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md).
+dans le [Guide Recherche, PDF, Excel & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md).

 Démarrage rapide :

@@ -234,7 +234,7 @@ Voir [PWA & mode hors-ligne](./PWA_HORS_LIGNE.md).

 | Objectif | Guide |
 |---|---|
-| Mieux chercher, lire PDF et Excalidraw | [Recherche, PDF & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md) |
+| Mieux chercher, lire PDF/Excel et Excalidraw | [Recherche, PDF, Excel & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md) |
 | Utiliser l'IA intégrée | [Assistant IA & Forge](./ASSISTANT_IA_FORGE.md) |
 | Éditer à plusieurs | [Édition & collaboration](./COLLABORATION.md) |
 | Sécuriser l'accès | [Authentification & sécurité](./AUTHENTIFICATION_SECURITE.md) |
@@ -15,7 +15,7 @@ captures conceptuelles).
 | Guide | Public | Contenu |
 |---|---|---|
 | 🚀 [Prise en main](./PRISE_EN_MAIN.md) | Tous | Premier lancement, interface, navigation, vaults, raccourcis |
-| 🔍 [Recherche, PDF & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md) | Tous | Syntaxe de requête, recherche sémantique, lecteur PDF, diagrammes |
+| 🔍 [Recherche, PDF, Excel & Excalidraw](./RECHERCHE_PDF_EXCALIDRAW.md) | Tous | Syntaxe de requête, recherche sémantique, lecteurs PDF/Excel, diagrammes |
 | 🤖 [Assistant IA & Forge](./ASSISTANT_IA_FORGE.md) | Tous | Fournisseurs, éditeur IA, BooksLM, Forge, commandes `@` / `/` |
 | 📝 [Édition & collaboration](./COLLABORATION.md) | Tous | Édition simultanée, curseurs distants, persistance |
 | 📱 [PWA & mode hors-ligne](./PWA_HORS_LIGNE.md) | Tous | Installation PWA, cache, file de synchronisation, notifications |
@@ -1,4 +1,4 @@
-# 🔍 Guide Recherche, PDF & Excalidraw
+# 🔍 Guide Recherche, PDF, Excel & Excalidraw

 ObsiGate va au-delà de la simple lecture : recherche puissante, rendu des
 documents riches (PDF, diagrammes) et indexation de leur contenu pour que tout
@@ -131,7 +131,83 @@ curl "http://localhost:2020/api/file/Recettes/pdf/info?path=menu.pdf"

 ---

-## 6. Diagrammes Excalidraw
+## 6. Tableurs Excel (XLSX)
+
+### Affichage et édition
+
+Un fichier `.xlsx` s'ouvre dans une visionneuse dédiée : un tableau par
+feuille, des onglets pour naviguer entre elles, les en-têtes A1/B1 et les
+numéros de ligne. Chaque cellule est modifiable directement (clic), `Entrée`
+valide, `Échap` annule la saisie. **Enregistrer** envoie les cellules
+modifiées à `PUT /api/file/{vault}/xlsx/save` : une sauvegarde par feuille,
+avec **backup automatique** du fichier avant écriture, et une écriture
+**atomique** (le classeur n'est jamais laissé à moitié écrit).
+
+### Avertissement avant enregistrement
+
+Certains classeurs contiennent des éléments qu'ObsiGate ne sait pas
+réécrire : **valeurs calculées** mises en cache par Excel, segments
+(slicers), chronologies, contrôles de formulaire, connexions/requêtes,
+XML personnalisé, signature numérique, commentaires enrichis, macros.
+L'ouverture affiche alors un bandeau qui les liste, et la première
+sauvegarde demande confirmation. Si vous refusez, rien n'est écrit.
+
+> Les **graphiques, images et tableaux croisés** sont, eux, bien conservés.
+
+### Formules
+
+Par sécurité, une valeur saisie commençant par `=` ou `@` est **stockée comme
+texte** (une formule injectée s'exécuterait à l'ouverture du fichier dans
+Excel). Le bouton `f(x)` de la barre d'outils active les vraies formules pour
+la session en cours.
+
+```bash
+curl -X PUT "http://localhost:2020/api/file/Recettes/xlsx/save?path=budget.xlsx"   -H "Content-Type: application/json"   -d '{"sheet": "Budget", "cells": {"B1": "250"}, "allow_formula": false, "force": false}'
+```
+
+- `allow_formula` : `true` pour écrire une vraie formule (`=B1*2`).
+- `force` : `true` pour enregistrer malgré les éléments non préservés
+  (sinon l'API répond **409** `xlsx_lossy_content`).
+- Deux sauvegardes simultanées sur le même fichier : la seconde reçoit
+  **409** `conflict` au lieu d'écraser la première.
+
+### Feuilles volumineuses et lecture par fenêtres
+
+Le rendu est plafonné à **500 lignes × 40 colonnes** par feuille. Quand
+une feuille dépasse ce plafond, un bandeau **« Feuille tronquée »**
+l'annonce explicitement (par exemple « 500 lignes affichées sur 520 »)
+au lieu de présenter une table courte comme complète — le classeur,
+lui, n'est jamais modifié. La ligne d'en-têtes de colonnes reste
+visible pendant le défilement vertical.
+
+Côté API, `GET /api/file/{vault}/xlsx/sheet` sert une feuille **par
+fenêtres de lignes**, y compris au-delà du plafond d'affichage — les
+coordonnées A1 renvoyées sont celles de la feuille réelle :
+
+```bash
+curl "http://localhost:2020/api/file/Recettes/xlsx/sheet?path=budget.xlsx&sheet=Budget&offset=500&limit=200"
+```
+
+- `offset` : première ligne renvoyée (0-based) ; `limit` : nombre de
+  lignes (1 à 1 000 par requête).
+- La réponse porte `total_rows`, `truncated` et `has_more` pour paginer.
+- Erreurs : **404** si la feuille n'existe pas, **415** si le fichier
+  n'est pas un `.xlsx`.
+
+### Limites
+
+- L'affichage intégré démarre à **500 lignes × 40 colonnes** par feuille ;
+  sous une feuille plus grande, le bouton **« Charger la suite »** (ou le
+  défilement vers le bas du tableau) ajoute les lignes suivantes par
+  fenêtres de 500 — elles deviennent aussitôt éditables et
+  sauvegardables.
+- Styles, formats de nombre, cellules fusionnées et volets figés ne sont pas
+  rendus.
+- Formats non gérés : `.xls`, `.xlsm` (macros), `.ods`.
+
+---
+
+## 7. Diagrammes Excalidraw

 Les fichiers `.excalidraw` et `.excalidraw.md` (dont le format compressé du
 **plugin Obsidian Excalidraw**) s'ouvrent dans un **éditeur visuel Excalidraw
@@ -147,7 +223,7 @@ Fiche technique : [`features/excalidraw.md`](../features/excalidraw.md).

 ---

-## 7. Autres contenus riches
+## 8. Autres contenus riches

 ### Mermaid

@@ -182,7 +258,7 @@ curl -X POST "http://localhost:2020/api/attachments/rescan/Recettes"

 ---

-## 8. Dépannage
+## 9. Dépannage

 | Symptôme | Piste |
 |---|---|
@@ -14,7 +14,7 @@

 - **Projet** : ObsiGate — Porte d'entrée web pour vaults Obsidian
 - **Stack** : Python 3.11+ (backend FastAPI) · JavaScript/Vanilla (frontend) · Tauri/Rust (desktop)
- **Dernière mise à jour** : 2026-09-17
+- **Dernière mise à jour** : 2026-09-27

 ---

@@ -181,13 +181,27 @@ Avant de corriger quoi que ce soit, un agent IA doit :
 | *BUG-070* | Activation clé physique WebAuthn impossible : « Validation du credential WebAuthn échouée » à chaque tentative | 🟢 corrigé | P0 | ⚙️ backend | IA | `backend/auth/webauthn_mfa.py`, `backend/auth/router.py` | Config → Sécurité → Ajouter une clé → cérémonie navigateur → 400 | `resolve_relying_party()` (rp_id/origines dérivés de la requête, config explicite prioritaire, forwarded si TRUST_PROXY) sur les 4 endpoints ; challenges multiples (5 derniers) acceptés ; `.env.example` ; tests `tests/test_webauthn.py` (+8) | Logs : origin `http://localhost:2020` rejetée + challenge mismatch au retry. Vérifié navigateur (authentificateur virtuel CDP) : register 200 + clé listée, clé de test retirée (admin de nouveau TOTP seul) ; pytest 1249 passed, ruff/mypy 0 |
 | *BUG-071* | Configuration « Configurations » inutilisable en mode mobile : sommaire masqué sans bouton d'accès, navigation par ancre sans JS, grilles 2 colonnes et rangées d'ajout qui débordent (≤768px) | 🟢 corrigé | P1 | 📱 frontend | IA | `frontend/index.html`, `frontend/js/config.js`, `frontend/js/i18n.js`, `frontend/style.css`, `frontend/locales/{fr,en}.json` | Mobile (≤768px) : ouvrir Configurations → aucun sommaire ni moyen d'atteindre une section ; champs « Clés IA » / jetons / webhooks débordent | `index.html` (+`#config-hamburger` `.help-hamburger`, `config.toc_toggle` FR/EN) ; `config.js` (toggle, scroll doux + actif + repli auto mobile, reset à l'ouverture) ; `i18n.js` (`data-i18n-attr` multi-paires `;`) ; `style.css` (bloc mobile `#config-modal` : sommaire haut 46vh, grilles 1fr, add-rows wrap + `!important`, items wrap, 44px) ; tests `tests/frontend/config-mobile.test.mjs` (nouveau, 11) + CI ; E2E `tests/e2e/config-mobile.spec.js` (nouveau, 3/3 projet chromium-mobile, ignoré en desktop) | pytest 1249 passed / 6 skipped, ruff 0, mypy 0, validate-imports 39 modules, unit 10/10, JSDOM ai 93/93 + sidebar 6/6 + mobile 35/35 + ai-keys 7/7 |
 | *BUG-072* | Visionneuse d'images : le plein écran et le panneau « Métadonnées » ne sont pas conservés lors de la navigation ←/→, et le panneau s'affiche sous la pellicule au lieu d'une barre latérale | 🟢 corrigé | P2 | 📱 frontend | IA | `frontend/js/viewer.js`, `frontend/style.css` | Ouvrir une image, activer le plein écran (ou Métadonnées), puis naviguer avec les flèches précédent/suivant | État persistant `_imageViewerState { lightbox, meta }` + drapeau `_imageViewerNavPending` posé par `go()`/pellicule : `renderFile` ne réinitialise que hors navigation image→image. Panneau reconstruit dans `.image-viewer-body` (sidebar droite, `border-left`, `width:280px; max-width:40%`) ; la règle lightbox ne masque plus que la pellicule. Boutons `image-btn-lightbox`/`image-btn-metadata` (+ `aria-pressed`), `Escape` resynchronisé. Tests : `tests/frontend/image-viewer.test.mjs` (+2), E2E `tests/e2e/image-viewer.spec.js` (+1). | Navigation → `openFile` → `renderImageViewer` recréait le conteneur : les états `lightbox`/`metaPanel` étaient perdus. Le panneau était rendu en bas (colonne) au lieu d'une sidebar droite |
+| *BUG-073* | Mobile : la barre de navigation fixe du bas masque le bas de tous les documents et pages affichés (les dernières lignes restent définitivement sous la barre) | 🟢 corrigé | P1 | 📱 frontend | IA | `frontend/style.css` | Mobile (≤768px) : ouvrir un document long, défiler jusqu'au bas → la fin du contenu passe sous la barre `#mobile-toolbar` et n'est jamais atteignable | Clearance mobile retargetée de `.main-layout` (sélecteur mort, absent de `index.html`) vers `.main-body` (`calc(64px + env(safe-area-inset-bottom, 0))`) ; règle sœur morte `.editor-modal.active ~ .main-layout` supprimée ; reset `body.reading-mode .main-body { padding-bottom: 0 }` (barre masquée en mode lecture) ; `body.np-active .content-area` ramené à `76px` (dégagement dock seul, géométrie totale inchangée). Tests : `tests/frontend/mobile-toolbar.test.mjs` (7, au CI), E2E `tests/e2e/mobile-toolbar.spec.js` (2, `chromium-mobile`, skip desktop) | La règle de clearance du bloc mobile cible `.main-layout`, classe absente de `index.html` (vrai conteneur : `.main-body`) → sélecteur mort, aucun dégagement réservé |
 |  | | | | | | | | | | |
+| *BUG-074* | [🟡 IMPORTANT] Assistant IA : le bloc d'étapes affiche « 1 step » sans titre alors que l'agent réalise plusieurs actions (compteur toujours à 1) | 🟢 corrigé | P1 | 📱 frontend + ⚙️ backend | IA | `frontend/js/bookslm.js`, `backend/agent/loop.py` | Mode agent : demander une création multi-fichiers/dossiers → chaque message ne montre qu'« 1 étape ▶ » sans détail | `backend/agent/loop.py` + `frontend/js/bookslm.js` : compteur = actions (hors réflexions) + titre = 1re action dans le `<summary>` ; reprise de confirmation diffusée dans le **même** message (fusion des étapes). Tests : `tests/frontend/ai.test.mjs` (+3) | Le résumé `<summary>` ne porte aucun titre ; chaque reprise de confirmation crée un **nouveau** message assistant qui ne contient qu'une action ; les réflexions gonflent le compteur |
+| *BUG-075* | [🟡 IMPORTANT] Assistant IA : chaque action mutatrice demande son propre « Appliquer » — aucun résumé des actions en attente ni approbation globale | 🟢 corrigé | P1 | ⚙️ backend + 📱 frontend | IA | `backend/agent/loop.py`, `backend/bookslm_routes.py`, `frontend/js/bookslm.js` | Mode agent : demander une structure de répertoires multi-fichiers → valider une action après l'autre | `backend/agent/loop.py` (`pending.actions`, lot exécuté au resume) ; `backend/bookslm_routes.py` (`confirm_all` → `ctx.confirmed`) ; `frontend/js/bookslm.js` (carte multi-actions + « Tout approuver (N) »). Tests : `tests/test_agent_loop.py`, `tests/test_bookslm.py`, `tests/frontend/ai.test.mjs` | La pause de confirmation ne capture que le **premier** appel mutateur du lot (les suivants sont `deferred`) ; carte unique sans liste ; nouveau `confirm_all` à ajouter pour autoriser la suite de l'exécution en une approbation |
+| *BUG-076* | [🟡 IMPORTANT] Assistant IA : après une action de l'agent, l'arborescence et le document ouvert ne sont pas rafraîchis dynamiquement | 🟢 corrigé | P1 | 📱 frontend | IA | `frontend/js/bookslm.js` | Mode agent : créer/supprimer un fichier ou dossier, modifier le document ouvert → l'UI ne bouge pas | `frontend/js/bookslm.js` : `MUTATING_TOOLS`/`FILE_WRITE_TOOLS`, refresh d'arborescence débouncé sur event `tool`, `_notifyFileWritten` étendu (xlsx/docx/csv/pdf). Tests : `tests/frontend/ai.test.mjs`, `tests/frontend/editor-inline.test.mjs` | Aucun refresh explicite sur les événements `tool` mutateurs (repose uniquement sur le watcher SSE) ; `_notifyFileWritten` ignore les créations de documents (xlsx/docx/csv/pdf) |
+| *BUG-077* | [🟡 IMPORTANT] Assistant IA : aucun bouton « Stop » pour arrêter l'exécution de l'agent à tout moment | 🟢 corrigé | P1 | 📱 frontend | IA | `frontend/js/bookslm.js` | Mode agent : lancer une longue tâche → le bouton Envoyer est désactivé, impossible d'arrêter (seule la fermeture du panneau abort) | `frontend/js/bookslm.js` + `frontend/style.css` : bouton d'envoi → Stop (`_syncSendButton`/`_stopGeneration`/`_markStopped`), i18n `ai.stop`/`ai.stopped`. Tests : `tests/frontend/ai.test.mjs` (+2) | `_abortCtrl` n'est déclenché que par `close()` ; aucun signal d'arrêt côté client pendant le stream |
+| *BUG-078* | [🟡 IMPORTANT] Fichiers de code : la coloration syntaxique (highlight.js) disparaît — les feuilles de thème sont basculées à partir de la **clé** de thème au lieu du **mode** | 🟢 corrigé | P1 | 📱 frontend | IA | `frontend/js/themes.js`, `frontend/js/ui.js`, `tests/frontend/unit.test.mjs` | Ouvrir un fichier `.py`/`.sh`/`.ps1`/`.yml` : le code s'affiche en texte brut, sans couleurs | `frontend/js/themes.js` : `applyTheme` bascule `hljs-theme-dark`/`hljs-theme-light` selon le **mode** (`isDark`). `frontend/js/ui.js` : `initTheme`/`applyTheme` résolvent le mode persisté (`obsigate-theme-mode`) au lieu de traiter la clé (`defaut-obsigate`) comme un mode. Test : `unit.test.mjs` (+1). | Les deux feuilles étaient désactivées car `defaut-obsigate !== "dark"` et `!== "light"` ; résultat **non déterministe** selon l'ordre `UI.initTheme()` (clé) / `Sync.init()` → `themes.initThemes()` (mode). Vérifié Playwright : 5/5 chargements colorés (`.py`), sépia/contraste élevé sur la palette claire |
+| *BUG-081* | `GET /api/auth/mfa/status` → 500 quand l'auth est désactivée (`user` None, `AttributeError` sur `user.get`) | 🟢 corrigé | P1 | ⚙️ backend | IA | `backend/auth/router.py::mfa_status`, `tests/test_mfa.py` | Auth désactivée : `curl http://127.0.0.1:2029/api/auth/mfa/status` → 500 (reproduit live 2026-09-27) | Garde `user is None` → payload MFA désactivé (`mfa_enabled: false`, `totp_enabled: false`, `webauthn_credentials: 0`) ; test `TestMfaStatusAuthDisabled` (échoue en 500 sans le correctif). Vérifié : `test_mfa.py` 32 passed, ruff/mypy 0 | `require_auth` laisse passer le pseudo-user anonymous, `get_user(username)` → None non gardé. Trouvé via les logs E2E pendant BUG-080 |
+| *BUG-079* | `GET /api/diagnostics` → 500 « dictionary changed size during iteration » (stats d'index) | 🟢 corrigé | P1 | ⚙️ backend | IA | `backend/main.py` | Charger la page de diagnostic pendant une indexation : `GET /api/diagnostics` → 500 | `backend/main.py` (`api_diagnostics`) : snapshot avant itération — `list(index.items())` et `inv.word_index.copy()` (copie C atomique sous le GIL) ; test de non-régression `tests/test_api_main.py::TestConfig::test_diagnostics_concurrent_index_writes` | Le handler itérait les dicts en direct alors que l'indexeur les modifiait depuis un autre thread (rebuild initial dans `_search_executor`, hooks incrémentaux `add_document`/`remove_document`) → `RuntimeError` dans le générateur → 500. Test déterministe (`RaceDict` fait grossir le dict en cours d'itération) : échoue sans le correctif, passe avec. Vérifié : pytest 1305 passed / 6 skipped, ruff 0, mypy 0 |
+| *BUG-084* | Index inversé : la suppression d'une vault y laisse des documents fantômes (résultats pour une vault inexistante) | 🟢 corrigé | P1 | ⚙️ backend | IA | `backend/indexer.py::remove_vault_from_index`, `backend/search.py::_remove_doc_internals` | Supprimer une vault configurée, puis chercher un terme contenu dans ses fichiers → les résultats la concernent encore | `remove_vault_from_index()` déclenche `_on_index_change('remove', …)` pour chaque fichier de la vault ; `_remove_doc_internals()` supprime la clé `vault_docs` dont le set devient vide (`defaultdict` : une lecture la recréait). Test `tests/test_search_advanced.py::TestVaultRemovalPurgesInvertedIndex` (contre-preuve : échoue sans le correctif) | Trouvé pendant la relecture de `plan.md` (étape 6 déjà livrée). Mesuré : 8 documents fantômes sur 8 après suppression de la vault de test (`postings`, `doc_info`, `doc_vault`, `vault_docs`) ; seul un reindex manuel les effaçait. Vérifié : `test_search_advanced.py` 27 passed, ruff/mypy 0, suite complète 1374 passed / 6 skipped |
+| *BUG-085* | Édition d'un `.xlsx` : les valeurs calculées en cache disparaissent du classeur (et tout lecteur `data_only=True` voit `None`) | 🟢 corrigé | P1 | tableur Excel | IA | `backend/xlsx_reader.py::inspect_workbook`, `backend/services/mutations.py::edit_xlsx_cells`, `backend/routers/files_read.py`, `backend/routers/files_write.py`, `frontend/js/viewer.js::renderXlsxViewer` | Ouvrir un classeur contenant `=B1*2` (avec sa valeur calculée) → éditer une cellule → le `<v>` disparaît du XML de la feuille | `LOSSY_PARTS` + sonde `<f>…</f><v>[^<]` ; la lecture renvoie `xlsx_lossy_features` ; `PUT xlsx/save` refuse sans `force` (**409** `xlsx_lossy_content`) ; bandeau + confirmation UI puis reprise `force: true`. Tests : `TestXlsxLossyGuard` (5) + `xlsx-viewer.test.mjs` (10) + `tests/e2e/xlsx-viewer.spec.js` (3) | #153 A1. Périmètre réel vérifié sur openpyxl 3.1.5 : graphiques, images, dessins **et** TCD survivent au round-trip ; les pertes sont valeurs en cache, slicers/chronologies, contrôles de formulaire, connexions/requêtes, custom XML, signature, commentaires enrichis, macros. Vérifié : `test_xlsx_viewer.py` 31 passed, suite 1390 passed / 6 skipped, ruff/mypy 0, E2E 3/3 |
+| *BUG-086* | Édition d'un `.xlsx` : `wb.save()` écrit en place, un plantage laisse un classeur corrompu | 🟢 corrigé | P1 | tableur Excel | IA | `backend/services/mutations.py::edit_xlsx_cells` | Simuler un `OSError` pendant `Workbook.save` → le fichier d'origine est tronqué | Écriture atomique : `wb.save(<nom>.<pid>.tmp)` puis `os.replace()` ; `.tmp` supprimé sur échec ; le backup `.bak` reste inchangé. Test : `TestXlsxAtomicWrite::test_failed_save_keeps_the_original` (octets identiques après échec) + `test_no_tmp_left_after_a_successful_save` | #153 A2. Le fichier temporaire a un suffixe `.tmp` → ignoré par le watcher (`_is_relevant` ne retient que les extensions supportées). Vérifié : cf. BUG-085 |
+| *BUG-087* | Édition d'un `.xlsx` concurrente (deux onglets, agent IA + viewer) : read-modify-write sans verrou, le dernier écrivain gagne silencieusement | 🟢 corrigé | P1 | tableur Excel | IA | `backend/services/mutations.py::_xlsx_write_lock` | Deux `PUT xlsx/save` simultanés sur le même fichier → une écriture est écrasée sans trace | Verrou par chemin (registre + garde, timeout 15 s) autour du cycle load → edit → `os.replace` ; attente dépassée → **409** `conflict`. L'endpoint est devenu `def` (sync) pour que l'attente s'exécute dans le threadpool et ne bloque pas la boucle d'événements. Test : `TestXlsxWriteLock` (2) | #153 A3. Verrou en mémoire, par processus : protège les cas d'un même serveur (le cas desktop/Tauri). Vérifié : cf. BUG-085 |
+| *BUG-088* | Injection de formule dans un `.xlsx` : une saisie `=cmd\|'/c calc'!A1` est stockée comme formule et s'exécute à l'ouverture dans Excel (DDE) | 🟢 corrigé | P0 | tableur Excel / sécurité | IA | `backend/services/mutations.py::_write_cell`, `backend/routers/files_write.py`, `frontend/js/viewer.js::renderXlsxViewer` | `PUT /api/file/V/xlsx/save` avec `{"sheet": "S", "cells": {"A1": "=1+1"}}` → la cellule sort en `data_type == "f"` | `cell.data_type = "s"` après affectation : le texte est stocké comme chaîne, aucun `<f>` n'est écrit. Opt-in via `allow_formula: true` (endpoint) et le bouton `f(x)` de la visionneuse (session, jamais persisté). Test : `TestXlsxFormulaGuard` (4) + `xlsx-viewer.test.mjs` (toggle) | #153 A4. `+`/`-` ne sont pas neutralisés : ils sont déjà convertis en nombre par `_coerce_xlsx_value`. Le handler global `ServiceError` expose désormais `code` + `details` (le client en a besoin pour le 409), et `api()` (frontend) les propage sur l'Error. Vérifié : cf. BUG-085 |
+| *BUG-089* | Un reindex manuel ne reconstruisait pas l'index inversé : la recherche TF-IDF continuait de servir un index périmé | 🟢 corrigé | P1 | ⚙️ backend / recherche | IA | `backend/indexer.py::reload_index`, `backend/indexer.py::reload_single_vault`, `backend/search.py` | Modifier le contenu d'un fichier, puis `GET /api/index/reload` → la recherche renvoie encore l'ancien contenu (ou rien pour un fichier nouveau) | `reload_index()` / `reload_single_vault()` appellent `init_inverted_index()` après le rebuild (le remplacement wholesale d'une entrée de vault n'émet pas les notifications incrémentales). En prime, `backend/search.py` lisait l'index via `from backend.indexer import index` (liaison **par valeur** du dict) : un `importlib.reload(backend.indexer)` recréait le dict côté indexer tandis que la recherche écrivait encore dans l'ancien — l'index inversé n'indexait alors plus rien. Tous les accès passent désormais par `_indexer.index`. Contre-preuve : `TestXlsxSearchable::test_search_finds_a_word_stored_in_a_cell` échoue sans le correctif | #153 A5. Trouvé en écrivant le test de recherche d'A5 : il passait isolément et échouait en suite complète selon l'ordre. Le reload incrémental par fichier (watcher, edition) n'est pas concerné : il passe par le hook `_on_index_change`. Vérifié : suite 1402 passed / 6 skipped, ruff/mypy 0 |
+| *BUG-090* | Troncature silencieuse d'une feuille `.xlsx` au-delà de 500 lignes × 40 colonnes : l'utilisateur voit une table courte sans aucun indice que la suite existe | 🟢 corrigé | P1 | tableur Excel / UX | IA | `backend/xlsx_reader.py::render_sheets`, `backend/routers/files_read.py`, `frontend/js/viewer.js::renderXlsxViewer`, `frontend/style.css` | Ouvrir `test_vault/sample-xlsx-large.xlsx` (520 lignes) → la feuille s'arrête à la ligne 500 sans aucun message | `render_sheets()` renvoie désormais `total_rows`/`total_cols` (dimensions déclarées par la feuille), `max_rows`/`max_cols` (plafonds du moteur) et `truncated` ; la visionneuse affiche un bandeau « Feuille tronquée — 500 lignes affichées sur 520 » (i18n `xlsx.truncated_*` FR/EN, axe des colonnes inclus). Contre-preuve : neutraliser `truncated` → `TestXlsxTruncationNotice` (2 tests) échoue | #153 A8/R5. La ligne d'en-têtes est aussi `sticky` au défilement vertical (`thead th { top: 0 }` + `top: auto` sur les numéros de ligne pour éviter l'empilement en haut à gauche). L'endpoint `GET …/xlsx/sheet` (#153 A9) sert les fenêtres au-delà du plafond, mais le chargement paresseux complet (défilement virtuel, « charger tout ») reste à faire — le bandeau dit la vérité en attendant. Vérifié : `test_xlsx_viewer.py` 58 passed, E2E 7/7 (dont 3 nouveaux), suite 1417 passed / 6 skipped, ruff/mypy 0, i18n parity |

 ### TODOs techniques (améliorations / nouvelles tâches)

 | # | Titre | Statut | Priorité | Scope | Assigné | Zone (fichier) | Cmd de repro | Correctif / Commit | Notes |
 |---|---|---|---|---|---|---|---|---|---|
-| *(exemple)* TODO-002 | Rendre l'index inversé incrémental (40k+ fichiers) | 🔴 ouvert | P1 | ⚙️ backend | IA | `backend/indexer.py`, `backend/search.py` | Recherche sur très gros vault | — | Exemple à remplacer. Cf. plan.md |
 | *(À remplir)* | | | | | | | | | |

 ---
@@ -200,6 +214,9 @@ Avant de corriger quoi que ce soit, un agent IA doit :

 | Date | ID(s) traité(s) | Action | Fichiers modifiés | Résumé | Statut après |
 |---|---|---|---|---|---|
+| 2026-09-28 | BUG-090 (#153 A8 + A9) | Correction + feature | `backend/xlsx_reader.py`, `backend/routers/files_read.py`, `backend/schemas.py`, `backend/openapi_docs.py`, `frontend/js/viewer.js`, `frontend/style.css`, `frontend/locales/{fr,en}.json`, `tests/test_xlsx_viewer.py`, `tests/frontend/xlsx-viewer.test.mjs`, `tests/e2e/xlsx-viewer.spec.js`, `test_vault/sample-xlsx-large.xlsx` | **La troncature d'une feuille est annoncée et les lignes cachées restent accessibles** : (BUG-090/A8) `render_sheets()` renvoie `total_rows`/`total_cols`/`max_rows`/`max_cols`/`truncated`, la visionneuse affiche un bandeau « Feuille tronquée » (i18n FR/EN, axes lignes et colonnes) et la ligne d'en-têtes devient `sticky` (`top: auto` sur les numéros de ligne pour éviter l'empilement) ; (A9) `GET /api/file/{vault}/xlsx/sheet?sheet=&offset=&limit=` (`XlsxSheetWindowResponse`, plafond 1 000 lignes/requête, 404 feuille inconnue, 415 non-xlsx) sert une fenêtre avec les **vraies** coordonnées A1 et le `has_more` de pagination. Contre-preuves : neutraliser `truncated` → 2 tests échouent ; neutraliser l'offset → 3 tests échouent. Vérifié : `test_xlsx_viewer.py` 58 passed, xlsx-viewer.test.mjs 14/14, E2E 7/7 (3 nouveaux + fixture `sample-xlsx-large.xlsx` 520 lignes), suite 1417 passed / 6 skipped, ruff 0, mypy 0, i18n parity, validate-imports 40 modules | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-28 | BUG-089 (#153 A5, A10, A12) | Correction | `backend/xlsx_reader.py`, `backend/indexer.py`, `backend/search.py`, `backend/services/mutations.py`, `frontend/js/viewer.js`, `frontend/style.css`, `frontend/locales/{fr,en}.json`, `tests/test_xlsx_viewer.py` | **Les tableurs deviennent visibles ettypés** : (A5) `extract_indexable_text()` indexe noms de feuilles + 20 premières lignes (plafond 5 k caractères) dans le TF-IDF et la recherche sémantique — un mot tapé dans une cellule rend le fichier trouvable ; (A10) `_coerce_xlsx_value()` reconnaît désormais les booléens (`TRUE`/`FAUX`/`OUI`/`NON`) et les dates FR `JJ/MM/AAAA` (jour-first : `01/02/2026` = 1er février), symétrique avec l'affichage ; (A12) la valeur calculée en cache s'affiche sous la formule (`<span class="xlsx-cached">`, 2ᵉ lecture `data_only=True` uniquement si l'archive contient un `<v>`), info-bulle traduite via `xlsx.cached_value_title` FR/EN. (BUG-089) un reindex manuel reconstruisait mal l'index inversé et `backend/search.py` lisait l'index par valeur. Contre-preuves vérifiées pour A5, A10 et A12. Vérifié : `test_xlsx_viewer.py` 43 passed, suite 1402 passed / 6 skipped, ruff 0, mypy 0, i18n parity, validate-imports 40 modules, xlsx-viewer.test.mjs 10/10 | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-27 | BUG-085 → BUG-088 (#153 A1-A4) | Correction | `backend/xlsx_reader.py`, `backend/services/mutations.py`, `backend/routers/files_read.py`, `backend/routers/files_write.py`, `backend/schemas.py`, `backend/main.py`, `frontend/js/viewer.js`, `frontend/js/auth.js`, `frontend/style.css`, `frontend/locales/{fr,en}.json`, `frontend/sw.js`, `tests/test_xlsx_viewer.py`, `tests/frontend/xlsx-viewer.test.mjs`, `tests/e2e/xlsx-viewer.spec.js`, `test_vault/sample-xlsx-lossy.xlsx`, `.gitea/workflows/ci.yml` | **Garde-fous d'écriture des classeurs Excel** : (BUG-085) `inspect_workbook()` détecte ce qu'un round-trip openpyxl perd (valeurs calculées, slicers, contrôles, connexions, custom XML, signature) → la lecture expose `xlsx_lossy_features`, la visionneuse affiche une bannière et `PUT xlsx/save` refuse sans `force` (**409** `xlsx_lossy_content`, confirmation explicite puis reprise) ; (BUG-086) écriture atomique `.tmp` + `os.replace` ; (BUG-087) verrou par fichier (409 `conflict`, endpoint sync pour le threadpool) ; (BUG-088) une saisie `=`/`@` est stockée en texte (`data_type = "s"`), sauf opt-in `allow_formula` / bouton `f(x)`. Le handler `ServiceError` expose désormais `code` + `details` et `api()` les propage. Périmètre de perte revalidé empiriquement sur openpyxl 3.1.5 (graphiques, images et TCD sont préservés). Vérifié : `test_xlsx_viewer.py` 31 passed, suite 1390 passed / 6 skipped, ruff/mypy 0, validate-imports 40 modules, xlsx-viewer.test.mjs 10/10, E2E 3/3 | 🟢 corrigé (en attente vérif utilisateur) |
 | *(exemple)* 2026-06-15 | BUG-001 | Correction | `frontend/app.js` | Réécriture de `renderFile()` pour préserver le DOM dashboard | 🟢 corrigé (en attente vérif) |
 | 2026-09-09 | BUG-001, BUG-002 | Correction | `backend/main.py`, `frontend/excalidraw-editor.html`, `tests/test_pdf_stream.py` | BUG-001: Content-Disposition RFC 5987 (nom PDF accentué ne casse plus l'en-tête → plus de 500). BUG-002: suppression alias esm.sh (408 jotai) + React 19 cohérent + prop `excalidrawAPI` → Loading masqué, save OK. Vérifié: 534 tests backend verts + E2E navigateur. | 🟢 corrigé (en attente vérif utilisateur) |
 | 2026-09-11 | BUG-003, BUG-004 | Correction | `backend/{main,indexer,export,pdf_reader,bookslm_routes}.py`, `backend/auth/router.py`, `.gitea/workflows/ci.yml`, `README.md`, `README.fr.md` | BUG-003: 33 erreurs mypy corrigées (annotations, gardes `None`, import `PROVIDERS` manquant → bug latent) + étape CI mypy rendue bloquante. BUG-004: lien `README.md` → `docs/CONTRIBUTING.md`. Vérifié: mypy 0 erreur, ruff OK, pytest 728 passed, frontend OK. | 🟢 corrigé (en attente vérif utilisateur) |
@@ -260,6 +277,17 @@ Avant de corriger quoi que ce soit, un agent IA doit :
 | 2026-09-23 | BUG-071 | Correction | `frontend/index.html`, `frontend/js/config.js`, `frontend/js/i18n.js`, `frontend/style.css`, `frontend/locales/{fr,en}.json`, `tests/frontend/config-mobile.test.mjs` (nouveau), `.gitea/workflows/ci.yml`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-071** : page « Configurations » inutilisable en mobile. (1) `#config-nav` masquée sous 768px sans toggle → hamburger `#config-hamburger` ajouté à l'en-tête (`.help-hamburger`, libellé `config.toc_toggle` FR/EN). (2) Ancres brutes sans JS → interception en `config.js` (scroll doux, lien actif, repli auto mobile, reset à l'ouverture). (3) Débordements 360px → bloc CSS mobile `#config-modal` (sommaire haut 46vh, grilles 1fr, add-rows wrap + largeurs inline neutralisées, items wrap, cibles 44px). `data-i18n-attr` multi-paires (`;`). Vérifié : `config-mobile.test.mjs` 11/11 (nouveau, au CI), pytest 1249 passed / 6 skipped, ruff/mypy 0, validate-imports 39 modules, unit 10/10, JSDOM ai 93/93 + ai-sidebar 6/6 + sidebar-filters 8/8 + mobile-editor 35/35 + config-ai-keys 7/7. | 🟢 corrigé (en attente vérif utilisateur) |
 | 2026-09-23 | BUG-071 (complément E2E) | Test | `tests/e2e/config-mobile.spec.js` (nouveau), `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-071 (complément E2E)** : spec Playwright mobile (convention `mobile-editor.spec.js` : `test.skip` hors viewport ≤768px, donc inactive sur le projet `chromium-desktop` du CI). Vérifié en local sur l'instance de test (port 2029, auth désactivée) : hamburger → sommaire, sélection → scroll + actif + repli, 0 débordement horizontal à 393px (3/3 `chromium-mobile`, 3 ignorés en desktop) ; suite `mobile-editor.spec.js` intacte (3/3). | 🟢 corrigé (en attente vérif utilisateur) |
 | 2026-09-23 | BUG-072 | Correction | `frontend/js/viewer.js`, `frontend/style.css`, `tests/frontend/image-viewer.test.mjs`, `tests/e2e/image-viewer.spec.js`, `scripts/run-e2e-local.ps1` (nouveau), `package.json`, `AGENTS.md`, `README.md`, `README.fr.md`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-072** : dans la visionneuse d'images (#108-D), le plein écran (lightbox) et le panneau « Métadonnées » étaient perdus dès qu'on changeait d'image avec ←/→ (ou la pellicule), car `openFile` → `renderFile` recrée entièrement `renderImageViewer`. (1) **Persistance** : état module `_imageViewerState { lightbox, meta }` restauré à chaque rendu ; un drapeau `_imageViewerNavPending` posé par `go()` et le clic de vignette indique à `renderFile` que le rendu suivant est une navigation image→image (pas de réinitialisation) — toute autre ouverture repart à zéro. (2) **Panneau latéral** : `.image-meta-panel` déplacé dans un nouveau `.image-viewer-body` en flex row, à droite de `.image-stage` (`border-left`, `width:280px; max-width:40%`, défilement vertical) au lieu d'une bande sous la pellicule ; la règle lightbox ne masque plus que la pellicule. Boutons stables `image-btn-lightbox`/`image-btn-metadata` + `aria-pressed`, `Escape` resynchronise l'état. Tests statiques `image-viewer.test.mjs` (+2) et E2E Playwright (+1). **Diagnostic E2E** : `npm run test:e2e` bloquait car `bash` résout vers WSL (HS, Ubuntu `Stopped`, `HCS_E_CONNECTION_TIMEOUT`) et git-bash est bloqué par App Control → lanceur PowerShell ajouté. Vérifié : `image-viewer.spec.js` 4/4, **suite `chromium-desktop` complète 103 passed / 6 skipped (10,3 min)** via `scripts/run-e2e-local.ps1`, `image-viewer.test.mjs` 12/12, unit 10/10, validate-imports 40 modules. | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-23 | BUG-073 | Correction | `frontend/style.css`, `tests/frontend/mobile-toolbar.test.mjs` (nouveau), `tests/e2e/mobile-toolbar.spec.js` (nouveau), `.gitea/workflows/ci.yml`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-073** : en mobile (≤768px), la barre fixe `#mobile-toolbar` (64px + safe-area) recouvrait le bas de tous les documents/pages — fin de contenu inaccessible. (1) **Cause** : la règle de clearance du bloc `@media (max-width: 768px)` ciblait `.main-layout`, classe absente de `index.html` (vrai conteneur : `.main-body`) → sélecteur mort, zéro dégagement ; règle sœur morte `.editor-modal.active ~ .main-layout { padding-bottom: 0 }` supprimée (overlay plein écran / nécessaire en édition inline). (2) **Correctif** : `.main-body { padding-bottom: calc(64px + env(safe-area-inset-bottom, 0)) }`, reset `body.reading-mode .main-body { padding-bottom: 0 }` (barre masquée en mode lecture), `body.np-active .content-area` ramené de `calc(64px+safe+76px)` à `76px` (dégagement dock seul — géométrie totale identique, pas de double comptage avec `.main-body`). Tests : `mobile-toolbar.test.mjs` (7 statiques, ajouté au CI), E2E `mobile-toolbar.spec.js` (géométrie + scroll fin de `ANALYSE_REVIEW.md`, skip hors viewport ≤768). Vérifié : `mobile-toolbar` 7/7, JSDOM 14 suites 0 échec, **E2E `chromium-mobile` BUG-073 2/2 + régressions mobile-editor/config-mobile 6/6**, **suite `chromium-desktop` complète 106 passed / 9 skipped (11,4 min)**, pytest 1302 passed / 6 skipped, ruff/mypy 0, validate-imports 40 modules, unit 11/11. | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-24 | #114 | Feature | `frontend/index.html`, `frontend/js/config.js`, `frontend/style.css`, `frontend/locales/{fr,en}.json`, `tests/frontend/config-mobile.test.mjs`, `tests/e2e/config-mobile.spec.js`, `docs/features/settings-mobile-114.md` (nouvelle), `docs/ROADMAP.md`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **#114 — Configuration, refonte mobile-responsive (≤768px)**. (1) **Drawer sommaire** : `#config-nav` en panneau `position: fixed` (`min(320px, 88vw)`, z-index 40) sous backdrop `#config-modal.config-toc-open::before` (z-index 35) ; bouton `#config-toc-close` ; backdrop/Échap ferment le drawer d'abord puis la modale ; `_setConfigNav` bascule la classe + conserve le `display` inline de reset. (2) **Modale plein écran** : `100dvh` + `padding: 0`, `border-radius: 0`. (3) **Tactile** : boutons/liens ≥44px, inputs/selects `16px` + `min-height: 44px` (anti-zoom iOS, scopé `#config-modal`), rangée `.config-actions-row` sticky column + safe-area, formulaires 1 colonne, MFA 1 colonne + code full-width, wrap webhook/token/share/diag/avatar/webauthn. (4) **Dettes HTML/i18n** : `.config-actions-row` replacée dans `#cfg-backend-settings` (`</section>` orphelin supprimé), id dupliqué `cfg-partages-publics` retiré du `<h2>`, `#plugins-settings-container` supprimé, doublons `.config-btn-add` + règle morte `.mfa-recovery-input` purgés, `#mt-explorer` → `data-i18n="settings.explorer"` ; i18n : clés mortes `settings.{backend,backend_hint,restart_badge,save,plugins}` supprimées, `settings.explorer` + `config.toc_close` ajoutées, `settings.tabs` FR = « Onglets ». Vérifié : `config-mobile.test.mjs` 27/27 (au CI), unit 11/11, validate-imports 40 modules, pytest 1302 passed / 6 skipped, ruff/mypy 0, E2E `chromium-mobile` 5/5. | ✅ livré (en attente vérif utilisateur) |
+| 2026-09-24 | BUG-074 → BUG-077 | Correction | `backend/agent/loop.py`, `backend/bookslm_routes.py`, `frontend/js/bookslm.js`, `frontend/style.css`, `frontend/index.html`, `frontend/locales/{fr,en}.json`, `frontend/sw.js`, `tests/test_agent_loop.py`, `tests/test_bookslm.py`, `tests/frontend/ai.test.mjs`, `tests/frontend/editor-inline.test.mjs`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **Lot assistant IA (mode agent)** : (BUG-075) confirmation par lot — `pending.actions` regroupe toutes les mutations d'un tour LLM, un unique bouton « Tout approuver (N) » envoie `confirm_all` (`ToolContext.confirmed`) et n'interrompt plus à chaque action ; les lectures du lot s'exécutent aussitôt. (BUG-074) résumé du bloc d'étapes avec titre de la 1re action + compteur limité aux actions, reprise diffusée dans le même message (fini le « 1 étape » fragmenté). (BUG-076) refresh de l'arborescence débouncé sur les events `tool` mutateurs + `_notifyFileWritten` étendu aux documents (xlsx/docx/csv/pdf). (BUG-077) le bouton d'envoi devient « Stop » pendant le stream (abort SSE, tâche serveur annulée à la déconnexion, marqueur « Exécution arrêtée. »). Guide/i18n FR/EN + `ai.stop`/`ai.stopped`/`ai.confirm_actions`/`ai.action_apply_all` ; `SW_VERSION` v25. Vérifié : pytest 1304 passed / 6 skipped, ruff/mypy 0, validate-imports 40 modules, unit 11/11, ai 100/100, editor-inline 44/44, mobile-editor 35/35, ai-sidebar 6/6, forge 32/32, pane-manager 9/9, sw 8/8. | 🟢 corrigé (en attente vérif utilisateur) |
+
+| 2026-09-24 | #115, #117, BUG-078 | Feature + correction | `frontend/js/themes.js`, `frontend/js/ui.js`, `frontend/js/viewer.js`, `frontend/js/config.js`, `frontend/index.html`, `frontend/style.css`, `frontend/popout.html`, `frontend/locales/{fr,en}.json`, `frontend/icons/avatar/*` (nouveau), `tests/frontend/unit.test.mjs`, `tests/frontend/toolbar-order.test.mjs`, `tests/frontend/settings-order-avatar.test.mjs`, `docs/features/viewer-toolbar-highlight-avatars.md` (nouvelle), `docs/ROADMAP.md`, `CHANGELOG.md` | **#115** barre d'outils de lecture épinglée : `viewer.js`/`popout.html` sortent `.file-actions` de `.file-header` dans un `.file-toolbar` enfant direct de `.content-area` (`position: sticky; top: 0`), masqué en mode lecture. **BUG-078** coloration syntaxique : le basculement des feuilles highlight.js suit le **mode** (`themes.applyTheme` + `ui.initTheme/applyTheme` lisent `obsigate-theme-mode`) au lieu de la clé de thème qui désactivait les deux feuilles. **#117** avatars prédéfinis : galerie de 12 images (`frontend/icons/avatar/`) dans `#cfg-profile`, clic → recadrage 256 px (pipeline import) + `PATCH /api/auth/me`, avatars actifs surlignés (`obsigate-avatar-preset`), import personnalisé et suppression conservés. Vérifié : Playwright (coloration 5/5 déterministe, toolbar épinglée à `barTop` constant au défilement), `unit.test.mjs` 12/12, `toolbar-order` 13/13, `settings-order-avatar` 12/12, JSDOM editor-inline/pane-manager/mobile-editor/image-viewer/pdf-viewer/config-mobile/media-viewer/excalidraw verts, pytest 1304 passed / 6 skipped, ruff/mypy 0, validate-imports 40 modules. | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-24 | BUG-079 | Correction | `backend/main.py`, `tests/test_api_main.py`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-079** : `GET /api/diagnostics` renvoyait 500 « dictionary changed size during iteration ». Le handler itérait `inv.word_index.values()` et `index.items()` en direct alors que l'indexeur les modifiait depuis un autre thread (rebuild initial dans `_search_executor`, hooks incrémentaux `add_document`/`remove_document`) → `RuntimeError` dans le générateur. Correctif : **snapshot avant itération** (`list(index.items())`, `inv.word_index.copy()`) — copie C atomique sous le GIL, pas de verrou ajouté. Test de non-régression déterministe (`RaceDict` fait grossir le dict pendant l'itération ; échoue sans le correctif, passe avec). Vérifié : pytest 1305 passed / 6 skipped, ruff 0, mypy 0 (80 fichiers), validate-imports 40 modules, unit 12/12. | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-27 | BUG-080, BUG-081 | Correction + enregistrement | `scripts/run-e2e-local.ps1`, `scripts/run-e2e-local.sh`, `scripts/e2e-server.ps1`, `playwright.config.ts`, `tests/test_e2e_harness.py` (nouveau), `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-080** : run E2E local pendu toute la nuit → harnais anti-blocage : `npx --yes` (plus de prompt interactif), install Chromium sautée si présent (`E2E_INSTALL_BROWSERS=1`), timeouts `E2E_TIMEOUT_SEC` (900)/`E2E_BROWSER_INSTALL_TIMEOUT_SEC` (600, exit 124), `globalTimeout` Playwright (15 min local / 30 min CI, `E2E_GLOBAL_TIMEOUT_MS`), pidfile resynchronisé sur le vrai owner du port + `stop` qui tue l'arbre complet (orphelins 81180/81936 nettoyés, port 2029 libéré). Diagnostic : double processus systématique (parent `.venv` parqué + enfant qui sert — environnemental, aussi sur flowdeck/3.13). **BUG-081** (ouvert, non traité) : `GET /api/auth/mfa/status` → 500 auth désactivée (`user` None, `router.py:827`, reproduit live). Vérifié : `test_e2e_harness.py` 8/8, cycle start/stop live (pidfile cohérent, port libéré). | 🟢 corrigé (en attente vérif utilisateur) ; BUG-081 🔴 ouvert |
+| 2026-09-27 | BUG-082 | Correction CI | `.gitea/workflows/ci.yml`, `tests/test_ci_workflow.py`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-082** : `lint` rouge (`ERR_MODULE_NOT_FOUND: jsdom`, rouge depuis `7bee4a2`) — les fichiers de l'étape frontend racine à import statique `jsdom` (`upload.test.mjs`, puis `config-ai-keys.test.mjs` révélé par le CI après le 1er fix), alors que `jsdom` n'est installé que dans `tests/frontend/node_modules` (étape JSDOM). Les deux déplacés dans l'étape JSDOM (les deux branches) ; garde-fou `test_ci_workflow.py` généralisé (aucun fichier racine à import statique jsdom + suites verrouillées en JSDOM, contre-preuve OK). Vérifié : étape racine verte (11 suites) + `upload` et `config-ai-keys` verts depuis `tests/frontend/`. | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-27 | BUG-083 | Correction CI | `.gitea/workflows/ci.yml`, `tests/test_ci_workflow.py` (nouveau), `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-083** : job `security` rouge — le runner Gitea Act tronque naïvement au premier `#` (même entre guillemets) : `echo "... see #87)"` devenait une citation non fermée (`unexpected EOF while looking for matching '"'"`, `/var/run/act/workflow/4` ligne 2). Seul `run:` du workflow avec un `#` (les `#` des noms d'étapes Bandit/Npm audit sont inoffensifs, ces étapes passent). Correctif : echo sans `#` (réf `#87` en commentaire YAML). Garde-fou `test_ci_workflow.py` (aucun `#` dans le code des `run:`, `upload.test.mjs` verrouillé en étape JSDOM — BUG-082) + contre-preuve sur l'ancien `ci.yml`. Vérifié : 56 passed. | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-27 | BUG-081 | Correction | `backend/auth/router.py`, `tests/test_mfa.py`, `CHANGELOG.md`, `docs/ISSUES_TODOLIST.md` | **BUG-081** : `GET /api/auth/mfa/status` répondait 500 quand l'auth est désactivée — le pseudo-user `anonymous` n'a aucune entrée en store (`get_user` → `None`, `AttributeError` sur `user.get`). Garde `user is None` → payload « MFA désactivé ». Test `TestMfaStatusAuthDisabled` (échoue en 500 sans le correctif). Vérifié : `test_mfa.py` 32 passed, ruff/mypy 0. | 🟢 corrigé (en attente vérif utilisateur) |
+| 2026-09-27 | #87 T6, T7, T8 | Sécurité (fin #87) | `backend/requirements.txt`, `backend/{render,export}.py`, `backend/tools/documents.py`, `backend/auth/router.py`, `backend/main.py`, `semgrep-rules/` (nouveau), `.gitea/workflows/ci.yml`, `tests/test_i18n_parity.py` (nouveau), `tests/test_auth_api.py`, `tests/test_security_headers.py`, `docker-compose.yml`, `.env.example`, `CHANGELOG.md`, `docs/ROADMAP.md`, `docs/ISSUES_TODOLIST.md` | **T6** : dépendances qualifiées (mistune 3.3.3, multipart 0.0.31, weasyprint 70, mcp 1.28.1, fastapi 0.141.1 + starlette 1.7.0, setuptools 84 ; `cast` mistune 3 sites) — suite 1359 passed, ruff/mypy 0, **`pip-audit` bloquant 0 vuln** (exception ecdsa/Minerva documentée : sans fix, HS256 only). **T7** : **semgrep bloquant** local 8 règles, 0 finding (trivy écarté : réseau). **T8** : Secure auto + `X-Forwarded-Proto` (`TRUST_PROXY`), warning affiné, CORS same-origin explicite, `style-src` résiduel assumé (189+343 sites) ; TODO exemple purgé, locales FR/EN 2213 parité testée, `npm audit` 0. | 🟢 corrigé (en attente vérif utilisateur) |

 ---

@@ -270,7 +298,9 @@ Avant de corriger quoi que ce soit, un agent IA doit :

 | # | Titre | Date résolution | Résolu par | Correctif / Commit | Notes |
 |---|---|---|---|---|---|
-| *(aucun pour l'instant)* | | | | | |
+| *BUG-083* | Job CI `security` rouge : le runner Gitea Act tronque le script `pip-audit` au premier `#` (citation de l'echo non fermée → `unexpected EOF while looking for matching '"'`) | 2026-09-27 | Utilisateur | `run:` assaini (echo sans `#`, réf `#87` en commentaire YAML) ; `tests/test_ci_workflow.py` (2 tests : aucun `#` dans le code des `run:`, `upload.test.mjs` verrouillé en étape JSDOM) ; vérifié : 56 passed (ci_workflow + e2e_harness + version), contre-preuve OK sur l'ancien `ci.yml` | Seul `run:` du workflow contenant un `#` (`see #87` dans l'echo). Les `#` des noms d'étapes (Bandit, Npm audit) sont inoffensifs (ces étapes passent). Correctif : echo sans `#`, réf `#87` en commentaire YAML |
+| *BUG-082* | CI `lint` rouge : suites frontend à import statique `jsdom` exécutées dans l'étape racine où `jsdom` n'est jamais installé | 2026-09-27 | Utilisateur | `upload.test.mjs` + `config-ai-keys.test.mjs` déplacés dans l'étape JSDOM (les deux branches) ; garde-fou `test_ci_workflow.py` (aucun fichier racine à import statique jsdom + suites verrouillées en JSDOM) ; vérifié : étape racine verte + `upload` et `config-ai-keys` verts depuis `tests/frontend/` | `jsdom` ne vit que dans `tests/frontend/node_modules` (installé par l'étape JSDOM). Correctif : déplacer les suites concernées dans l'étape JSDOM |
+| *BUG-080* | [🔴 BLOQUANT] E2E locaux bloqués toute la nuit : `npm run test:e2e:ps` ne termine jamais (serveurs orphelins sur le port 2029, `npx playwright install` sans `--yes` ni garde-fou, suite ~130 tests sans timeout global) | 2026-09-27 | Utilisateur | `run-e2e-local` : `npx --yes`, skip install Chromium si présent (`E2E_INSTALL_BROWSERS=1`), timeouts `E2E_TIMEOUT_SEC` (900)/`E2E_BROWSER_INSTALL_TIMEOUT_SEC` (600, exit 124) ; `playwright.config.ts` : `globalTimeout` 15 min local / 30 min CI (`E2E_GLOBAL_TIMEOUT_MS`) ; `e2e-server.ps1` : pidfile = vrai owner du port, `stop` tue l'arbre complet. Tests : `tests/test_e2e_harness.py` (8/8), cycle start/stop live (pidfile cohérent, port libéré) | Constat 2026-09-27 : `e2e-server.ps1 start` OK (READY 12 s) mais run suivant pendu toute la nuit ; 2 python orphelins (PID 81180 parent + 81936 sur le port, pidfile périmé). Double processus systématique (parent `.venv` parqué + enfant qui sert — aussi sur flowdeck/3.13 : environnemental, sans impact après correctif). Trouvé au passage : BUG-081 (`/api/auth/mfa/status` → 500 auth désactivée) |

 ---

@@ -1,6 +1,6 @@
 # ObsiGate — Roadmap

-> **Version :** 2.22.0 | **Dernière mise à jour :** 2026-09-23
+> **Version :** 2.33.0 | **Dernière mise à jour :** 2026-09-28
 > **Ce fichier ne contient que le travail à venir** (🔵 En cours + ⚪ Backlog) et un index compact
 > vers les fonctionnalités livrées.
 > - **Méthode de livraison à appliquer pour toute tâche : [DELIVERY_WORKFLOW.md](./DELIVERY_WORKFLOW.md)**
@@ -37,14 +37,67 @@
 - **Reste à faire :**
  - [x] **Signature de l'updater Tauri** (gratuit) : paire de clés générée, `pubkey` renseignée, `createUpdaterArtifacts` activé, secrets CI câblés
  - [x] **Manifeste `latest.json`** généré par `scripts/updater_manifest.py` (intégré à `publish_release.py`), endpoint updater pointé sur `main`
-  - [ ] **Signature de code Windows** : non retenue (pas de certificat) — alternatives : livrer non signé, SignPath.io (OSS gratuit), Certum OSS, Azure Trusted Signing, certificat EV
+  - [ ] **Signature de code Windows** : **non retenue — décision confirmée le 2026-09-26** : livraison non signée + documentation SmartScreen (« Exécuter quand même »). Alternatives écartées sauf retour utilisateur : SignPath.io (OSS gratuit), Certum OSS, Azure Trusted Signing, certificat EV
  - [ ] Exécuter les 6 tests E2E **manuels** — protocole documenté : [DESKTOP_E2E_CHECKLIST.md](./DESKTOP_E2E_CHECKLIST.md)

+---
+
+## 🔵 En cours — Visionneuse & édition Excel (P0/P1/P2)
+
+### 153. Visionneuse & édition XLSX — complétude (fidélité, recherche, IA, UX, formats)
+
+- **Effort :** 8-13 jours (P0 ✅ 2-3 j · P1 : 4-6 j · P2 : 2-4 j) | **Impact :** 🟡
+- **Statut :** 🔵 en cours — **P0 livré le 2026-09-27** (BUG-085 → BUG-088), **A5/A10/A12 livrés le 2026-09-28** (avec BUG-089), **A8/A9/A9bis livrés le 2026-09-28** (avec BUG-090), reste A6-A7 puis A13-A17
+- **Analyse, risques et critères d'acceptation :** [features/xlsx-viewer.md](./features/xlsx-viewer.md)
+- **Description :** #152 (visionneuse XLSX, 2.27.0) lit et édite correctement la **grille de
+  valeurs** d'un `.xlsx`, mais l'ensemble supporté est étroit : valeurs seulement (ni structure,
+  ni styles en écriture, ni formule recalculée), **écriture destructive** d'une partie du classeur,
+  tableurs **invisibles à la recherche** et **inutilisables par l'IA** au-delà de la création. Ce
+  lot suit ces ajouts ; les cases ci-dessous sont le **suivi de référence**, la fiche feature porte
+  le détail.
+- **Constat (points de départ) :** `MAX_ROWS = 500` / `MAX_COLS = 40` sans indicateur (troncature
+  silencieuse) · `wb.save()` non atomique et sans verrou (concurrence) · saisie `=…` stockée comme
+  formule par openpyxl (injection DDE) · `content=""` à l'indexation (recherche TF-IDF et sémantique
+  aveugles) · aucun outil IA de lecture/édition d'un classeur existant · aucun test frontend ni
+  E2E sur le viewer.
+- **Périmètre réel des pertes au round-trip (mesuré sur openpyxl 3.1.5, 2026-09-27) :** graphiques,
+  images, dessins **et** tableaux croisés sont préservés ; sont perdus les **valeurs calculées en
+  cache**, slicers/chronologies, contrôles de formulaire, connexions/requêtes, custom XML,
+  signature numérique, commentaires enrichis et macros.
+- **Sous-tâches :**
+  - **P0 — garde-fous d'écriture (🔴, 2-3 j) — 🟢 livré**
+    - [x] **A1** Alerte de fidélité avant écriture : `inspect_workbook()` → `xlsx_lossy_features` + bandeau FR/EN + **409** `xlsx_lossy_content` sans `force` (confirmation explicite puis reprise) — BUG-085
+    - [x] **A2** Écriture atomique (`wb.save(.tmp)` + `os.replace()`, backup inchangé) — BUG-086
+    - [x] **A3** Verrou par fichier autour du read-modify-write (timeout 15 s + **409** `conflict`) — BUG-087
+    - [x] **A4** Neutralisation de l'injection de formule (`=`/`@` stockés en texte, opt-in `allow_formula` + bouton `f(x)`) — BUG-088
+  - **P1 — recherche, IA, UX (🟡, 4-6 j) — 🔵 en cours**
+    - [x] **A5** Indexation du contenu des feuilles (noms de feuilles + 20 premières lignes, plafond 5 k caractères) — les mots tapés dans une cellule rendent le fichier trouvable ; au passage **BUG-089** (reindex manuel ne reconstruisait pas l'index inversé)
+    - [ ] **A6** Outils IA `update_xlsx_cells` / `append_xlsx_rows` / `xlsx_to_markdown` / `list_xlsx_sheets`
+    - [ ] **A7** Navigation clavier + barre de formule + nom de cellule (Tab/Entrée/flèches, `Maj+Entrée`, copie de plage)
+    - [x] **A8** `thead` sticky + bandeau « feuille tronquée » (lève la troncature silencieuse) — BUG-090
+    - [x] **A9** Chargement paresseux par feuille (`GET …/xlsx/sheet?offset&limit`, défilement virtuel)
+    - [x] **A10** Types & formats de saisie (nombre/texte, booléens `TRUE`/`FAUX`, dates FR `JJ/MM/AAAA` jour-first)
+    - [ ] **A11** Tests frontend (`tests/frontend/xlsx-viewer.test.mjs`) + E2E (`tests/e2e/xlsx-viewer.spec.js`) au CI
+    - [x] **A12** Valeur calculée affichée sous la formule (2ᵉ lecture `data_only=True` seulement si l'archive contient un `<v>`, info-bulle FR/EN)
+  - **P2 — étendu (🟢, 2-4 j) — ⚪ à faire**
+    - [ ] **A13** Tri / filtre / recherche dans la feuille + export CSV de la sélection
+    - [ ] **A14** CRUD de feuilles, lignes et colonnes (renommer, insérer, supprimer, dupliquer)
+    - [ ] **A15** Styles minimaux en écriture + lecture fidèle (gras, fond, formats, fusions, volets figés)
+    - [ ] **A16** Formats additionnels (`.xlsm` avec `keep_vba`, `.xls`, `.ods`, `.csv` éditable)
+    - [ ] **A17** Vue « tableau de bord » (plages nommées, TCD, KPI par feuille, actions IA)
+- **Convention de suivi :** chaque sous-tâche démarre par son ID stable (`#153-A<n>` dans cette
+  Roadmap) ; celles qui sont des **défauts** sont aussi ouvertes comme `BUG-NNN` dans
+  [ISSUES_TODOLIST.md](./ISSUES_TODOLIST.md) (A1→BUG-085, A2→BUG-086, A3→BUG-087, A4→BUG-088 ;
+  A8 le sera à son tour).
+
+---
+
 ## ⚪ Backlog — Priorité 4 (P4)

 ### 73. Synchronisation multi-appareils — Obsidian Sync compatible

 - **Effort :** 6-8 jours | **Impact :** 🟢
+- **Décision 2026-09-26 : reporté (P4)** — axe prioritaire = dette & sécurité (#85/#87) ; #73 hors chemin critique. Si réactivé : partir d'un MVP export/hash/LWW adossé à #59 (PWA offline) + #62 (collab Yjs/CRDT) plutôt qu'un protocole parallèle.
 - **Description :** Synchronisation des vaults entre plusieurs instances d'ObsiGate via un protocole de synchronisation décentralisé ou compatible Obsidian Sync. Alternative self-hosted à Obsidian Sync.
 - **Sous-tâches :**
  - [ ] Protocole : évaluation CRDT vs OT vs diff/patch pour fichiers markdown
@@ -58,60 +111,22 @@

 ---

-## ⚪ Backlog — Priorité 2 (P2)
-
-### 83. Barre d'outils d'édition mobile — style Obsidian Android
-
- **Effort :** 3-5 jours | **Impact :** 🟡 | **Zone :** frontend (mobile)
- **Statut :** ✅ livré — ruban horizontal défilable ancré au-dessus du clavier, commandes étendues et personnalisation persistée. Détail : [archive/COMPLETED_v1-v2.md](./archive/COMPLETED_v1-v2.md) (section #83).
- **Description :** remplacer la barre de mise en forme Markdown actuelle par un **ruban horizontal
-  défilable** ancré juste au-dessus du clavier virtuel, reprenant l'ergonomie de l'app Android
-  Obsidian : fond anthracite aux coins arrondis, insertion/enrobage de la syntaxe au curseur ou sur
-  la sélection, et personnalisation des commandes via une icône clé à molette.
- **Sous-tâches :**
-  - [x] Ruban horizontal défilable (glissement tactile gauche/droite) ancré au-dessus du clavier
-  - [x] Actions rapides : annuler, refaire, `[[ ]]` (lien interne), modèle/fichiers, tag `#`, pièce jointe
-  - [x] Formatage : H1–H6, gras, italique, barré (`~~`), surligné (`==`), code en ligne/bloc, citation (`>`)
-  - [x] Liens externes, listes à puces/numérotées, case à cocher (`- [ ]`), indenter / désindenter
-  - [x] Personnalisation (clé à molette) : ajouter / supprimer / réordonner les commandes
-  - [x] i18n FR/EN + tests frontend (helpers purs) + E2E mobile
-
---
-
 ## ⚪ Backlog — Sécurité, architecture & performance (P0/P1)

-### 84. Consolidation & sécurité — revue statique 2026-09-13 (phase 1)
-
- **Effort :** 6-9 jours | **Impact :** 🔴 | **Zone :** backend + frontend | **Référence :** [ISSUES_TODOLIST.md](./ISSUES_TODOLIST.md) BUG-021 → BUG-034
- **Statut :** 🟢 livré (phase 1) — sanitizer XSS, rate-limit/lockout MFA, isolation vaults, ReDoS, SSRF webhooks, cycle de vie des sessions, politique de mot de passe, verrous `users.json`, audits IP, rate-limit par compte, symlinks, recherche via inverted index, token en cookie HttpOnly. Détail : [archive/COMPLETED_v1-v2.md](./archive/COMPLETED_v1-v2.md) (section #84).
- **Description :** traiter toutes les vulnérabilités critiques et importantes issues de la revue statique : XSS markdown (`escape=False`) et page publique de partage, brute-force MFA, isolation des vaults (`resolve_safe_path`), ReDoS, SSRF webhooks, cycle de vie des sessions, politique de mot de passe, races `users.json`, audits IP, rate-limit partagé, indexation symlinks.
- **Sous-tâches :**
-  - [x] Assainir le rendu markdown (sanitizer serveur en whitelist) et la page de partage (échappement `title`/frontmatter) — *DOMPurify client non ajouté (défense en profondeur serveur suffisante)*
-  - [x] Rate-limit + lockout sur les endpoints MFA (`totp/verify`, `recovery`, `webauthn/verify`)
-  - [x] Corriger `resolve_safe_path` (comparaison de chemin stricte par segment) + test de régression
-  - [x] Rotation du refresh token, révocation de l'access token au logout, persistance des JTI révoqués
-  - [x] Valider la politique de mot de passe à la création ; bloquer le SSRF des webhooks et externaliser les secrets
-  - [x] Verrous sur les mutations `users.json` ; consigner l'adresse IP réelle dans les audits
-  - [x] Ignorer les symlinks de l'index ; caps CPU/timeout regex (ReDoS)
-  - [~] Durcir la CSP — *partiel* : directives `object-src`/`base-uri`/`form-action`/`frame-ancestors` ajoutées et token retiré de `sessionStorage` ; migration **nonce** restante (nécessite la conversion des gestionnaires d'événements inline)
-
-### 85. Refonte architecturale — découpage du monolithe & persistance d'état (phase 2)
-
- **Effort :** 8-12 jours | **Impact :** 🟡 | **Zone :** backend
- **Description :** extraire le monolithe `backend/main.py` (~4 260 lignes) en routers FastAPI par domaine et rendre persistant l'état qui ne l'est pas (index de recherche, JTI révoqués, compteurs de rate-limit) pour préparer le multi-nœuds.
- **Sous-tâches :**
-  - [ ] Routers par domaine : files, search, share, webhooks, plugins, collab, admin, ai
-  - [ ] Centraliser le contrat d'outils IA sur `tools/registry.py` (permissions, quotas, redaction)
-  - [ ] Persister index, JTI révoqués et compteurs de rate-limit (SQLite/Redis)
-  - [ ] Verrous asyncio autour de l'index global et des stores JSON ; service de partage public (expiration, révocation, quotas)
-
 ### 87. Amélioration continue — tests, CI/CD, revues de sécurité (phase 4)

 - **Effort :** 3-5 jours | **Impact :** 🟡 | **Zone :** `.gitea/workflows/`, `tests/`
- **Description :** renforcer le pipeline (`.gitea/workflows/ci.yml`, `desktop-build.yml`) pour le rendre bloquant par défaut et accompagner les phases 1 → 3.
+- **Décision 2026-09-26 : prioritaire (axe Dette & sécurité).**
+- **Statut :** 🔵 en cours depuis 2026-09-26 — par tranches. **T1 livrée (v2.28.1) :** bandit bloquant (`nosec` justifiés B324/B404/B603/B607/B406, B105 exclu comme `pyproject`), `npm audit` bloquant (0 vulnérabilité), 5 suites frontend intégrées au CI (`upload`, `pretty`, `media-viewer`, `mfa-settings`, `config-ai-keys`). pip-audit reste consultatif (montées starlette/weasyprint à qualifier).
+- **T6 livrée (v2.28.15) :** dépendances qualifiées — mistune 3.3.3, python-multipart 0.0.31, weasyprint 70, mcp 1.28.1, fastapi 0.141.1 + starlette 1.7.0, setuptools 84 (`cast` mistune 3 sites) — suite 1359 passed, ruff/mypy 0, **`pip-audit` bloquant, 0 vulnérabilité** (seule exception documentée : PYSEC-2026-1325 ecdsa, sans correctif upstream, JWT HS256 uniquement).
+- **T7 livrée (v2.28.15) :** **semgrep bloquant** sur ruleset 100 % local `semgrep-rules/` (8 règles, 0 finding, contrôle négatif OK) ; trivy écarté (binaire + DB réseau, couche Python couverte).
+- **T8 livrée (v2.28.15, fin BUG-034) :** cookies `Secure` auto (`true|false|auto`, `X-Forwarded-Proto` sous `TRUST_PROXY`, warning affiné, `TRUST_PROXY=true` en prod) ; `CORSMiddleware` same-origin explicite ; `style-src 'unsafe-inline'` conservé assumé (189 `style=` + 343 `el.style`, T5c ayant verrouillé `script-src`).
+- **Description :** renforcer le pipeline (`.gitea/workflows/ci.yml`, `desktop-build.yml`) pour le rendre bloquant par défaut et accompagner les phases 1 → 3. Constat 2026-09-26 : job `security` non bloquant (`bandit`/`pip-audit` en `|| echo`, ni semgrep ni trivy), E2E limité à `chromium-desktop`, 5 suites frontend hors CI.
 - **Sous-tâches :**
-  - [ ] Jobs CI sécurité (bandit/semgrep/trivy, audits pip/npm) + tests E2E XSS (page de partage + lecteur markdown)
-  - [ ] Tests de concurrence (`users.json`), fuzzing de timing regex, couverture des composants critiques
+  - [ ] Jobs CI sécurité **bloquants** (bandit/semgrep/trivy, audits pip/npm) + tests E2E XSS (page de partage + lecteur markdown) — **T4 livrée :** `tests/e2e/xss.spec.js` (BUG-021/022, 2/2 vert) + `scripts/e2e-server.ps1` (cycle de vie serveur E2E avec progression `start|stop|status|logs`) + validation locale projet `chromium-desktop` : **108/108 verts** (obsigate 44, split 37, viewers 24, xss/header 3), mobiles ciblés 10/10
+  - [ ] Tests de concurrence (`users.json`), fuzzing de timing regex, couverture des composants critiques ; intégrer au CI les 5 suites frontend hors CI (`upload`, `pretty`, `media-viewer`, `mfa-settings`, `config-ai-keys`) — **T2 livrée (v2.28.2) :** `tests/test_hardening_concurrency.py` (users.json concurrent + budget temps regex) ; 5 suites au CI (T1)
+  - [ ] Finir BUG-034 (migration CSP **nonce**, conversion des handlers inline), `Secure` cookies à `true` par défaut, politique CORS same-origin explicite ; confirmer la rotation de la clé DeepSeek (BUG-006, clé dans l'historique Git) — **T3 livrée (v2.28.3)** (helper + avertissement + CORS attesté) ; **T5a livrée (v2.28.6)** (16 handlers inline → listeners, CSP inchangée) ; **T5b livrée :** nonce frais par réponse (`backend/csp.py`, `script-src`), injection dans les 6 pages HTML (dont nouvelle route `/excalidraw-editor.html`), `unsafe-inline` conservé (inerte) ; **T5c livrée (v2.28.13)** (`script-src` sans `unsafe-inline`) ; **T8 livrée (v2.28.15)** (fin BUG-034 : Secure auto + CORS explicite ; `style-src` résiduel assumé ; rotation DeepSeek BUG-006 toujours côté utilisateur)
+  - [ ] Revue périodique des dépendances ; documentation utilisateur FR/EN synchronisée ; contrôle automatisé de la conformité au DoD — **T6/T9 livrées (v2.28.15)** (`pip-audit` 0, `npm audit` 0, locales FR/EN 2213 clés parité testée `test_i18n_parity.py`, gardes `test_version.py` + `test_ci_workflow.py`)
  - [ ] Revue périodique des dépendances ; documentation utilisateur FR/EN synchronisée ; contrôle automatisé de la conformité au DoD

 ---
@@ -124,6 +139,7 @@

 | # | Domaine / fonctionnalité | Version | Détails |
 |---|---|---|---|
+| 152 | Viewer XLSX — affichage multi-feuilles, édition des cellules, téléchargement | 2.27.0 | [archive](./archive/COMPLETED_v1-v2.md) |
 | BUG-047 | Versionnage — source unique `VERSION` + bump SemVer automatique au commit (hooks + tag) | 2.3.0 | [DEVELOPMENT_AND_RELEASES.md](./DEVELOPMENT_AND_RELEASES.md) |
 | 90 | Barre d'actions du document — regroupement fonctionnel + spacers | 2.3.0 | [archive](./archive/COMPLETED_v1-v2.md) |
 | 89 | Drag & drop complet de fichiers/dossiers & intégration Assistant IA | 2.3.0 | [features/drag-and-drop-ai.md](./features/drag-and-drop-ai.md) |
@@ -186,6 +202,11 @@
 | 111 | Visionneuse d'images — navigation fluide : image ajustée au cadre, navigation en place (cache annuaire + préchargement), pellicule persistante, flèches latérales au survol | 2.20.0 | [features/image-navigation-111.md](./features/image-navigation-111.md) |
 | 112 | En-tête allégé & compte en sidebar — version dans le menu Options, retrait utilisateur/déconnexion du header, section compte en bas de la sidebar, pellicule d'images défilable (molette + flèches) | 2.21.0 | [features/header-user-sidebar-112.md](./features/header-user-sidebar-112.md) |
 | 113 | Configuration — ordre naturel des sections (Profil 1er, À propos dernier, TOC = page) & avatar utilisateur (import PNG/JPG/WEBP, persistance serveur, cercle sidebar) | 2.22.0 | [features/settings-order-avatar-113.md](./features/settings-order-avatar-113.md) |
+| 114 | Configuration — refonte mobile-responsive (modale plein écran 100dvh, sommaire en drawer coulissant, cibles tactiles ≥ 44 px, inputs 16 px anti-zoom, sauvegarde sticky, MFA 1 colonne, purge i18n/HTML) | 2.23.0 | [features/settings-mobile-114.md](./features/settings-mobile-114.md) |
+| BUG-078 | Fichiers de code — coloration syntaxique restaurée (feuilles highlight.js basculées sur le mode de thème et non la clé) | 2.25.0 | [features/viewer-toolbar-highlight-avatars.md](./features/viewer-toolbar-highlight-avatars.md) |
+| 115 | Viewer — barre d'outils de lecture épinglée au défilement | 2.25.0 | [features/viewer-toolbar-highlight-avatars.md](./features/viewer-toolbar-highlight-avatars.md) |
+| 117 | Configuration — avatars prédéfinis dans le profil utilisateur (12 images) | 2.25.0 | [features/viewer-toolbar-highlight-avatars.md](./features/viewer-toolbar-highlight-avatars.md) |
+| 85 | Refonte architecturale — découpage du monolithe (14 routers, `main.py` 4 827 → ~750 lignes), stores JSON verrouillés, rate-limit SQLite optionnel | 2.27.2→2.27.13 | [features/archi-refonte-85.md](./features/archi-refonte-85.md) |

 ---

@@ -193,16 +214,20 @@

 | Priorité | Items | Effort total estimé |
 |---|---|---|
-| ✅ Complété | #1 → #59, #61–72, #74–76, #78–84, #86, #88–93, #94–100, #102–113, #92 | ~131 jours réalisés |
-| 🔵 P2 restant | #77 Desktop : signature de code (non retenue), 6 tests E2E **manuels** ([protocole](./DESKTOP_E2E_CHECKLIST.md)) | ~0,5-1 jour |
-| ⚪ P4 restant | #73 Sync (6-8j) | 6-8 jours |
-| ⚪ P0/P1 restant | #85, #87 Refonte architecturale, CI/CD (BUG-035 → BUG-040 corrigés, #86 livré) | ~11-17 jours |
-| **Total restant** | **6 items + finitions** | **~23-38 jours** |
+| ✅ Complété | #1 → #59, #61–72, #74–76, #78–86, #88–93, #94–100, #102–115, #117, #92 | ~141 jours réalisés |
+| 🔵 Finitions | #77 Desktop : 6 tests E2E **manuels** ([protocole](./DESKTOP_E2E_CHECKLIST.md)) — signature Windows non retenue (décision 2026-09-26) | ~0,5-1 jour |
+| ⚪ P4 reporté | #73 Sync — **reporté (décision 2026-09-26)**, hors chemin critique | 6-8 jours si réactivé |
+| ⚪ P0/P1 prioritaire | #87 CI/CD (BUG-035 → BUG-040 corrigés, #86 livré) | ~3-5 jours |
+| ⚪ P0/P1/P2 backlog | #153 Visionneuse & édition XLSX — complétude (P0 ✅ A1-A4 ; P1 ✅ A5, A8-A10, A12, A9bis — reste A6-A7 ; A13-A17 2-4 j) | 2-4 jours restants |
+| **Total chemin critique** | **#77 fin + #87** | **~4-6 jours** |

 ---

 ## Notes

+- **Décisions 2026-09-26 :** axe prioritaire = dette & sécurité (#85/#87) ; #73 Sync reporté (P4, hors chemin critique) ; desktop livré non signé + doc SmartScreen.
+- **Ajout 2026-09-27 :** #153 ouvert à la suite de l'audit de la visionneuse XLSX (limitations, risques de perte de données, périmètre IA/recherche) — détail et critères dans [features/xlsx-viewer.md](./features/xlsx-viewer.md).
+- **Clôture #85 (v2.27.13) :** monolithe découpé (T1→T9), stores verrouillés + rate-limit SQLite (T10), fiche `docs/features/archi-refonte-85.md`.
 - Les items P3/P4 ne sont pas ordonnés par priorité interne — à raffiner selon les retours utilisateurs.
 - L'effort inclut le développement + tests unitaires + intégration CI, mais pas la documentation utilisateur.
 - Les items marqués 🟢 (nice-to-have) sont de bons candidats pour des contributions externes.
@@ -404,6 +404,21 @@ Deux compléments au bouton « Ajouter » de l'assistant IA.

 ---

+## #152 — Viewer XLSX : affichage, édition, téléchargement ✅ TERMINÉ
+
+Les fichiers `.xlsx` s'ouvrent dans un dédié : un tableau HTML par feuille (onglets en cas de
+multi-feuilles, en-têtes A1, cellules `contenteditable`), bouton **Enregistrer** actif dès la
+première modification et téléchargement du fichier d'origine.
+
+| Aspect | Détail |
+|---|---|
+| Lecture | `backend/xlsx_reader.py` — openpyxl `read_only`, formules affichées comme texte, plafond 500×40 cellules par feuille |
+| Écriture | `PUT /api/file/{vault}/xlsx/save` → `services/mutations.edit_xlsx_cells` (backup avant écriture, refs A1 validées, `str`→`int`/`float`, 500 cellules max par requête) |
+| Frontend | `renderXlsxViewer` dans `frontend/js/viewer.js` (onglets, cellules sales, Entrée/Échap, collage monoligne) |
+| Limite connue | Le round-trip openpyxl conserve valeurs/formules/styles mais perd graphiques, images et tableaux croisés |
+
+---
+
 ## Grosses fonctionnalités — fiches dédiées

 | # | Feature | Version | Fiche |
@@ -20,6 +20,7 @@
  - [x] **B3.** Fallback : retry sans `tools` si le provider rejette les tools (400/404/422) → chat simple ; protocole texte `obsigate-action` conservé côté frontend **pour le chat classique uniquement** (BUG-053 : en mode agent, le prompt impose les outils natifs et interdit les blocs `obsigate-action`)
  - [x] **B4.** SSE réellement streaming — `ai_chat.stream_completion` (`_openai_stream` + `_gemini_stream`) alimente `/api/ai/bookslm/chat` token par token ; le middleware GZip laisse passer les endpoints SSE BooksLM.
  - [x] **B5.** Confirmations UI : toggle « mode agent » (front → `/agent`), événements `tool`/`confirmation`, carte Apply + aperçu diff (LCS) pour les mutations, reprise `confirm`/`confirm_messages` côté backend. *S'active dès que la phase D enregistre des outils `write`.*
+    - **Complément (BUG-074 → BUG-077)** : la pause de confirmation **regroupe toutes les mutations** d'un même tour LLM (`pending.actions`, chacune avec son libellé `step` et son diff) et la carte n'offre plus qu'un seul bouton « **Tout approuver (N)** » ; la reprise envoie `confirm_all` et le backend arme `ToolContext.confirmed` pour le reste du run (plus d'approbation action par action). Le bloc d'étapes affiche un **titre** (1re action) et ne compte que les **actions** (hors réflexions) ; la reprise diffuse dans le **même message** (« N étapes » cumulées). L'arborescence et le document affiché sont **rafraîchis** dès une action mutatrice (refresh débouncé + `obsigate:file-written`, documents xlsx/docx/csv/pdf inclus). Le bouton d'envoi devient « **Stop** » pendant le stream (abort SSE, tâche serveur annulée à la déconnexion, marqueur « Exécution arrêtée. »).
  - [x] **B6.** Outils de navigation in-app : `open_file`, `reveal_in_tree` (événement `obsigate:open-file`) — livré via les liens cliquables de l'assistant (#80, [ai-assistant-ux.md](./ai-assistant-ux.md))
  - [x] **B7.** Tests : agent loop LLM mocké (`tests/test_agent_loop.py`), providers (`tests/test_ai_chat.py`), endpoint (`tests/test_bookslm.py`)

@@ -0,0 +1,77 @@
+# #85 — Refonte architecturale : découpage du monolithe & persistance d'état (phase 2)
+
+> **Statut :** livré (T1→T10) — `backend/main.py` 4 827 → ~750 lignes, 14 routers,
+> persistance partielle (stores verrouillés + rate-limit SQLite optionnel).
+> Méthode : tranches à impact minimal, comportement inchangé, un domaine par
+> commit, suite complète verte à chaque commit (1320 passed / 6 skipped).
+
+## 1. Découpage du monolithe (T1→T9, comportement inchangé)
+
+Chaque tranche déplace un domaine vers `backend/routers/` (handlers verbatim,
+mêmes chemins/modèles/auth/tags OpenAPI), les modèles vers `backend/schemas.py`,
+et ne committe que sur suite verte + `test_version` vert.
+
+| Tranche | Domaine | Nouveau module | Version |
+|---|---|---|---|
+| T1 | health (`/api/health*`) | `routers/health.py` (+ `HealthResponse` → schemas) | 2.27.2 |
+| T2 | webhooks CRUD | `routers/webhooks.py` | 2.27.3 |
+| T3 | sharing (`/api/share*`, `/s/*`) | `routers/sharing.py` | 2.27.4 |
+| T4 | backups (9 routes) | `routers/backups.py` (+ `Diff/Restore*` → schemas, `backend/sse.py`) | 2.27.5 |
+| T5 | search (11 routes) | `routers/search.py` (+ modèles → schemas, `backend/search_executor.py`) | 2.27.6 |
+| T6a | lecture fichiers | `routers/files_read.py` (+ modèles, `routers/helpers.py`) | 2.27.7 |
+| T6b | mutations fichiers/dossiers | `routers/files_write.py` (+ 15 modèles → schemas) | 2.27.8 |
+| T6c | media/pdf/export/guide | `routers/files_media.py` (Range helper → `helpers.py`) | 2.27.9 |
+| T7 | config (12 routes) | `routers/config.py` (`_FALLBACK_MODELS` déplacé) | 2.27.10 |
+| T8 | vaults + history + conflicts (13 routes) | `routers/vaults.py`, `history.py`, `conflicts.py` (+ `backend/watcher_state.py`) | 2.27.11 |
+| T9 | realtime + render | `routers/realtime.py` (SSE + collab WS), `backend/render.py` | 2.27.12 |
+
+`main.py` ne contient plus que l'assemblage : lifespan, middlewares, montage
+des routers, racine `/api`, statique/SPA, 4 cales de compatibilité testées
+(`_resolve_safe_path`, `_backup_file`, `_check_vault_writable`, `_get_backup_dir`).
+
+Correctifs au passage : décorateur orphelin `/s/{token}` (double-enregistrement
+de `/api/conflicts`), tag OpenAPI `media` inexistant (assignation par chemin
+conservée), tests statiques frontend réalignés (`image-viewer`, `media-viewer`),
+tests repointés vers les modules canoniques (`test_ai_models`, `test_api_main`).
+
+## 2. Persistance d'état (T10)
+
+| État | Avant | Après |
+|---|---|---|
+| JTI révoqués (`revoked_tokens.json`) | persisté, **sans verrou** | `RLock` (load/save/revoke/check) |
+| `shares.json` | persisté, **sans verrou** | `RLock` (4 mutateurs) |
+| `webhooks.json` + secrets | persistés, **sans verrou** | `RLock` (create/update/delete/secrets) |
+| `api_keys.json` (tool-secrets) | persisté, **sans verrou** | `RLock` (set/delete) |
+| Rate-limit auth | mémoire, mono-process | **inchangé par défaut** + option `OBSIGATE_RATELIMIT_DB` (SQLite WAL : mêmes fenêtres/budgets, partagé multi-workers, survit au redémarrage) |
+| Index de recherche | mémoire, rebuild au démarrage | **conservé** (voir §3) |
+| `users.json`, `api_tokens.json`, `vault_settings.json` | déjà verrouillés (BUG-029, #107) | inchangé |
+
+Tests : `tests/test_store_locks.py` (4 — concurrence threads, pertes prouvées
+sans verrou : 25/200 partages), `tests/test_ratelimit_store.py` (7 —
+sémantique SQLite identique, persistance, concurrence 200/200).
+
+Déjà existants et vérifiés (pas de code) : verrous `threading` + `asyncio`
+de l'indexeur (`_index_lock`, `_async_index_lock`), contrat central des
+outils IA — `backend/tools/registry.py` couvre déjà permissions
+(`requires_vault`, `require_destructive_allowed`), quotas
+(`check_and_record` par outil) et redaction (`redact_payload`) pour les
+35 outils enregistrés via `@tool(`.
+
+## 3. Décisions assumées (non fait, et pourquoi)
+
+- **Index non persisté sur disque.** Le rebuild différentiel (#86 : réutilise
+  les entrées inchangées `size` + `mtime`) rend le démarrage rapide ; un
+  snapshot introduirait des risques de staleness/drift de format sans gain
+  mesuré. Réévaluer si le démarrage devient lent (vaults 50k+ fichiers).
+- **Redis exclu.** SQLite WAL couvre le multi-workers mono-hôte sans nouvelle
+  infra ; Redis reste l'option multi-nœuds documentée (cf. `ratelimit.py`).
+- **`.gitignore` (`_*.py` ignore les `__init__.py`).** Contourné par
+  `git add -f` comme les packages existants ; assainir la règle à part.
+- Noms en `_` conservés (`backend/render.py`, stores) : déplacement verbatim,
+  zéro churn d'appels.
+
+## 4. Reste connu (hors #85)
+
+- CSP `unsafe-inline` (migration nonce, BUG-034 partiel) et `Secure` cookies → #87.
+- `main.py` (~750 lignes) : lifespan, middlewares, statique/SPA — cible
+  d'extraction ultérieure si besoin, non bloquant.
@@ -0,0 +1,172 @@
+# #114 — Configuration — refonte mobile-responsive de la section Settings
+
+> **Statut :** 🟢 · **Impact :** 🟡 · **Zone :** frontend (mobile, ≤ 768 px)
+> **Fichiers :** `frontend/index.html`, `frontend/js/config.js`, `frontend/style.css`,
+> `frontend/locales/{fr,en}.json`, `tests/frontend/config-mobile.test.mjs`,
+> `tests/e2e/config-mobile.spec.js`.
+
+## Contexte
+
+La page **Configurations** (`#config-modal`) était utilisable en mobile seulement
+partiellement (correctifs BUG-071) : le sommaire s'ouvrait en bloc haut, la modale
+n'était pas plein écran, les cibles tactiles étaient sous 44 px, le clavier virtuel
+iOS zoomait les champs, la rangée « Sauvegarder » disparaissait au scroll et plusieurs
+dettes HTML/i18n étaient restées en place.
+
+## Ce qui a été livré
+
+### A. Sommaire en drawer coulissant
+
+- `#config-nav` devient un **panneau coulissant gauche** (`position: fixed`,
+  `width: min(320px, 88vw)`, `z-index: 40`) sous un fond assombri
+  (`#config-modal.config-toc-open::before`, `z-index: 35`).
+- Ouverture/fermeture pilotée par la classe **`.config-toc-open`** sur `#config-modal`
+  (JS `_setConfigNav`) + un `display` inline conservé pour le contrat de reset.
+- Bouton **`#config-toc-close`** (classe `.help-toc-close`, `aria-label`
+  `config.toc_close`) dans l'en-tête du drawer ; visible uniquement dans le drawer
+  de la config (masqué sur desktop où la nav est toujours visible).
+- **Backdrop** : un tap hors du drawer ferme d'abord le drawer, pas la modale
+  (`e.target === modal` → `config-toc-open` présent → `_setConfigNav(false)`).
+- **Échap** : ferme le drawer d'abord, puis la modale.
+- Fermeture de la modale (`closeConfigModal`) nettoie toujours la classe
+  `config-toc-open` et le `display` inline (aucun fond ne subsiste).
+- Animation `config-toc-slide-in` (translateX) à l'ouverture.
+
+### B. Modale plein écran
+
+- `#config-modal` : `padding: 0`, `.editor-container` en `100vw × 100dvh`
+  (`100dvh` = hauteur du viewport dynamique, tient compte de la barre du navigateur
+  mobile), `border-radius: 0`, `border: none`.
+- Le contenu (`#config-scroll`) garde son scroll propre ; le sous-bloc
+  `.config-content` reçoit un `padding-bottom: 80 px` pour ne jamais passer sous la
+  rangée sticky.
+
+### C. Cibles tactiles ≥ 44 px & anti-zoom iOS
+
+- **Champs** : `.config-input`, `.config-select`, `.help-nav-search`,
+  `.profile-field .config-input/.config-select` → `min-height: 44px` +
+  `font-size: 16px` (**anti-zoom iOS** : un `font-size < 16px` déclenche le zoom
+  automatique au focus). La règle est **scoppée `#config-modal`** pour ne pas écraser
+  `.mfa-code-input` (qui a sa propre typographie).
+- **Boutons** : `.config-btn-save`, `.config-btn-secondary`, `.config-btn-primary`,
+  `.config-btn-danger`, `.config-btn-add`, `.config-btn-sm`, `.mfa-link-btn`,
+  `.theme-action-btn`, `.profile-avatar-actions .config-btn-secondary`,
+  `.editor-btn` (fermer), `#config-hamburger`, `#config-toc-close`,
+  `.help-search-clear` → `min-height/min-width: 44px`.
+- **Liens du sommaire** : `.help-nav-link` → `min-height: 44px` (ligne tactile
+  confortable).
+- `.help-hamburger` passe de 36 px à **44 px** en mobile (règle partagée avec le
+  modal d'aide).
+
+### D. Rangée « Sauvegarder » sticky
+
+- `.config-actions-row` (dans `#cfg-backend-settings`) → `position: sticky;
+  bottom: 0`, empilée verticalement (`flex-direction: column`), boutons pleine
+  largeur 44 px, fond opaque + bordure, `padding-bottom` avec
+  `env(safe-area-inset-bottom)` (barre home iOS).
+- La rangée reste visible pendant le scroll de la section backend ; le
+  `padding-bottom: 80px` de `.config-content` garantit qu'elle ne masque jamais les
+  derniers contrôles.
+
+### E. Formulaires 1 colonne & grilles
+
+- `.config-row` → 1 colonne (`grid-template-columns: 1fr`), `.config-input--num`
+  pleine largeur, `text-align: left`.
+- `.ai-default-grid` / `.ai-provider-fields` → 1 colonne.
+- Add-rows (`.config-add-row`, `.config-add-pattern`) → wrap + largeurs inline
+  (`180/140/100px`) neutralisées (`width: auto !important`).
+- Items webhook/token/share → wrap ; URLs/méta sur leur propre ligne
+  (`overflow-wrap: anywhere`) ; boutons de suppression 44 px.
+- `.hidden-files-add-row` → wrap, input pleine largeur.
+- `.config-diag-row` → wrap.
+- `.profile-avatar-row` → wrap ; `.profile-form` → `max-width: 100%`.
+- `.webauthn-key-item` → wrap ; `.webauthn-key-label` → pleine largeur.
+- `.theme-grid` reste en `auto-fill minmax(160px, 1fr)` (déjà responsive).
+
+### F. MFA & sécurité
+
+- `.mfa-recovery-list` → **1 colonne** en mobile (2 colonnes illisibles à 360 px).
+- `.mfa-verify-section`, `.mfa-recovery-actions`, `.mfa-disable-actions` → wrap ;
+  champs/boutons enfants en pleine largeur.
+- `.mfa-code-input` → `width: 100%`, `max-width: 320px`, `letter-spacing: 6px`
+  (au lieu de 12 px qui débordait), `min-height: 52px`.
+- `.mfa-secret-code` → `word-break: break-all` (secret TOTP long).
+- `.mfa-code-input-group` → wrap.
+- Règle morte **`.mfa-recovery-input`** supprimée (auth.js utilise
+  `mfa-code-input recovery-input`).
+
+### G. Dettes HTML corrigées
+
+- `.config-actions-row` (Sauvegarder / Réindexer / Réinitialiser) déplacée
+  **dans** `#cfg-backend-settings` (elle était hors de toute section → le sticky
+  n'avait pas de conteneur de scroll fiable) ; `</section>` orphelin supprimé.
+- Id dupliqué **`cfg-partages-publics`** retiré du `<h2>` (l'id reste sur la
+  `<section>`, cf. #113).
+- Conteneur mort **`#plugins-settings-container`** supprimé (le rendu réel est
+  `#cfg-plugins` via `plugins.js`).
+- Sections plugins/about ré-indentées.
+- **`#mt-explorer`** : libellé brut `settings.search` remplacé par
+  `<span data-i18n="settings.explorer">` (i18n correct, FR « Explorateur » / EN
+  « Files »).
+- Doublons CSS **`.config-btn-add`** (3 définitions) réduits à la définition de
+  référence.
+
+### H. i18n FR/EN
+
+- **Purge des clés mortes** (aucune référence HTML/JS/tests/backend) :
+  `settings.backend`, `settings.backend_hint`, `settings.restart_badge`,
+  `settings.save`, `settings.plugins`.
+- **Ajouts** : `settings.explorer` (FR « Explorateur » / EN « Files »),
+  `config.toc_close` (FR « Fermer le sommaire » / EN « Close contents »).
+- **Correctif** : `settings.tabs` en FR était le mot anglais « Tabs » →
+  **« Onglets »** (EN reste « Tabs »).
+- Clés vivantes conservées : `settings.reindex`, `settings.no_restart_badge`,
+  `settings.backend_section`, `settings.security`, `settings.search`,
+  `settings.tabs`, `settings.search_placeholder`.
+
+## Tests
+
+### Statics — `tests/frontend/config-mobile.test.mjs` (27, au CI)
+
+- BUG-071a–e (hamburger, toggle JS, grilles/wrap, ancres mortes,
+  `data-i18n-attr` multi-paires) — conservés et adaptés au drawer.
+- **#114a** : `#config-toc-close` présent + i18n ; `_setConfigNav` bascule
+  `.config-toc-open` ; backdrop `::before` (z-index 35) ; `.help-toc-close`
+  masqué sur desktop / visible dans le drawer ; Échap et backdrop ferment le
+  drawer d'abord ; `closeConfigModal` nettoie la classe.
+- **#114b** : modale `100dvh` + `padding: 0` ; inputs/selects `16px` +
+  `44px` ; boutons `44px` ; rangée sticky (`position: sticky` +
+  `flex-direction: column` + safe-area) ; MFA 1 colonne + wrap + code
+  full-width ; `.config-actions-row` bien dans `#cfg-backend-settings`.
+- **#114i18n** : clés mortes purgées ; `settings.explorer` FR/EN ;
+  `settings.tabs` FR = « Onglets » ; `#mt-explorer` porte
+  `data-i18n="settings.explorer"` ; `#plugins-settings-container` absent.
+
+### E2E — `tests/e2e/config-mobile.spec.js` (5, projet `chromium-mobile`)
+
+1. Hamburger → drawer `position: fixed` + classe `config-toc-open` sur la modale.
+2. Bouton `#config-toc-close` et tap backdrop ferment le drawer **sans** fermer la
+   modale.
+3. Sélection d'une section → scroll doux + lien actif + repli du drawer.
+4. Modale plein écran (largeur/hauteur ≈ viewport) + `#config-hamburger`,
+   `#config-close` et `#cfg-save-backend` ≥ 44 px.
+5. Aucun débordement horizontal à 393 px (sections IA / tokens / webhooks /
+   partages).
+
+## Détails d'implémentation notables
+
+- **Spécificité CSS** : les règles `#config-modal #config-nav` (2 ids) priment sur
+  les règles génériques `.help-nav` / `body .help-nav` qui masquent la nav en
+  mobile — pas besoin de `!important`.
+- **Backdrop = pseudo-élément** : les clics sur `::before` sont attribués à
+  l'élément or (`#config-modal`), donc le handler `e.target === modal` existant
+  fonctionne sans node supplémentaire.
+- **Contrat de reset conservé** : `configNavOnOpen.style.display = ''` à
+  l'ouverture (test statique BUG-071b) — le CSS reprend la main (drawer masqué par
+  défaut sur mobile).
+- **`100dvh` avec repli `100vh`** : les navigateurs sans support `dvh` gardent le
+  comportement précédent.
+- La règle `.help-hamburger { display: inline-flex }` du bloc mobile du modal
+  d'aide est **partagée** (44 px) ; le drawer de la config double avec
+  `#config-modal .help-hamburger` pour rester robuste à un réordonnancement des
+  règles.
@@ -0,0 +1,122 @@
+# #115 / BUG-078 / #117 — Barre d'outils épinglée, coloration syntaxique des fichiers de code & avatars prédéfinis
+
+> **Statut :** 🟢 livré (en attente vérification utilisateur)
+> **Impact :** 🟡 (#115, BUG-078) · 🟢 (#117)
+> **Zone :** frontend (`frontend/js/viewer.js`, `frontend/js/themes.js`,
+> `frontend/js/ui.js`, `frontend/js/config.js`, `frontend/index.html`,
+> `frontend/style.css`, `frontend/popout.html`, `frontend/locales/fr.json`,
+> `frontend/locales/en.json`, `frontend/icons/avatar/*`)
+
+Trois demandes traitées dans la même livraison, toutes côté frontend.
+
+## #115 — Barre d'outils de lecture toujours visible
+
+### Problème
+
+La barre d'actions d'un document (pop-out, bookmark, Editer, Source, Copier, PDF, Export,
+Partager…) était rendue **dans** `.file-header`, un conteneur court placé en haut de la
+zone de lecture. Or un élément `position: sticky` reste borné par son parent : dès que
+`.file-header` sortait de l'écran au défilement d'un document long, la barre disparaissait.
+
+### Correctif
+
+- `frontend/js/viewer.js` et `frontend/popout.html` sortent la barre d'actions de
+  `.file-header` et la placent dans un nouveau conteneur `.file-toolbar`, **enfant direct
+  de `.content-area`** (le conteneur de défilement). La barre peut donc se coller au haut
+  de la zone de lecture pour toute la hauteur du document.
+- `frontend/style.css` :
+
+  ```css
+  .file-toolbar {
+    position: sticky;
+    top: 0;
+    z-index: 30;
+    margin: 0 0 16px;
+    padding: 8px 0;
+    background: var(--bg-primary);
+    border-bottom: 1px solid var(--border);
+  }
+  ```
+
+  Le fond est opaque pour qu'aucun texte ne transparaisse dessous.
+- `body.reading-mode .file-toolbar { display: none }` : la barre reste masquée en mode
+  lecture, comme les autres actions.
+
+## BUG-078 — Coloration syntaxique des fichiers de code
+
+### Problème
+
+Les fichiers `.py`, `.sh`, `.ps1`, `.yml`, `.json`… (et les blocs de code Markdown)
+s'affichaient en **texte brut**, sans couleurs. Le code était pourtant correctement
+généré côté backend (`<pre><code class="language-python">…`) et `safeHighlight()` appelait
+bien `hljs.highlightElement()`.
+
+La cause était ailleurs : les deux feuilles de style de highlight.js (`#hljs-theme-dark`
+et `#hljs-theme-light`, chargées depuis le CDN dans `index.html`) étaient basculées à
+partir de la **clé de thème** persistée (`obsigate-theme` = `defaut-obsigate`, …) :
+
+```js
+darkSheet.disabled = theme !== "dark";    // "defaut-obsigate" !== "dark" → true
+lightSheet.disabled = theme !== "light";  // "defaut-obsigate" !== "light" → true
+```
+
+Les deux feuilles finissaient donc désactivées, privant tous les tokens de leurs couleurs.
+Le résultat dépendait de l'ordre de deux initialisations concurrentes — `UI.initTheme()`
+(clé de thème) au démarrage et `themes.initThemes()` (mode) via `Sync.init()` — d'où un
+comportement **non déterministe** (parfois coloré, le plus souvent non).
+
+### Correctif
+
+- `frontend/js/themes.js` — `applyTheme(themeKey, mode)` bascule désormais les feuilles
+  highlight.js selon le **mode** :
+
+  ```js
+  var isDark = mode === 'dark';
+  darkSheet.disabled = !isDark;
+  lightSheet.disabled = isDark;
+  ```
+
+  (`sepia` et `high-contrast` réutilisent la palette claire.)
+- `frontend/js/ui.js` — `initTheme()` lit le **mode** persisté (`obsigate-theme-mode`) et
+  `applyTheme()` résout un mode avant de fixer `data-theme` et de basculer les feuilles,
+  au lieu de comparer la clé de thème à `"dark"`/`"light"`.
+
+Le basculement est ainsi **déterministe** dès le premier rendu et à chaque changement de
+mode.
+
+## #117 — Avatars prédéfinis dans le profil
+
+### Ce qui a été livré
+
+- **Galerie de 12 avatars** dans la section `Profil` (`#cfg-profile`), servis depuis
+  `frontend/icons/avatar/` (`/static/icons/avatar/<fichier>.jpg`) : Chat, chien, elephan,
+  hibou, koala, lapin, lion, ours, penda, pingouin, raton, tigre.
+- Un clic charge l'image, la fait passer par le **même pipeline que l'import** (recadrage
+  carré central, redimensionnement 256 px, export JPEG 0,85) et l'enregistre via
+  `PATCH /api/auth/me` — aucune modification backend n'a été nécessaire, la validation
+  data-URL PNG/JPEG/WebP existante (BUG-113) s'applique telle quelle.
+- L'avatar actif est **surligné** ; le choix est mémorisé dans `localStorage`
+  (`obsigate-avatar-preset`) et purgé dès qu'on importe une photo personnalisée ou qu'on
+  supprime l'avatar.
+- L'import d'une photo et la suppression restent disponibles.
+- i18n : nouvelle clé `config.avatar_presets_label` (FR/EN).
+
+## Tests
+
+- `tests/frontend/unit.test.mjs` — `syntax highlight theme` (BUG-078) : `themes.applyTheme`
+  bascule les feuilles selon le mode et `ui.js` ne compare plus la clé au mode.
+- `tests/frontend/toolbar-order.test.mjs` — #115 : `.file-toolbar` dans `viewer.js` et
+  `popout.html`, règle CSS `position: sticky; top: 0`, masquage en mode lecture.
+- `tests/frontend/settings-order-avatar.test.mjs` — #117 : 12 avatars présents dans
+  `#cfg-profile`, fichiers d'images existants, pipeline `config.js`, règles CSS, clé i18n
+  FR/EN.
+- Vérification Playwright (instance locale, auth désactivée) : coloration déterministe sur
+  5 chargements successifs d'un `.py` ; `.file-toolbar` dont le `top` ne bouge plus après
+  défilement (épinglage effectif).
+
+## Limitations
+
+- L'avatar reste stocké en data-URL 256 px dans `data/users.json` (comportement #113
+  conservé).
+- Le mode `high-contrast`/`sepia` utilise le thème highlight.js **clair** (pas de palette
+  dédiée).
@@ -0,0 +1,235 @@
+# #153 — Visionneuse & édition XLSX — état des lieux et backlog
+
+> **Item de roadmap :** [#153 — Visionneuse & édition XLSX — complétude](../ROADMAP.md)
+> **Origine :** #152 (visionneuse XLSX, livrée en 2.27.0 — voir
+> [archive/COMPLETED_v1-v2.md](../archive/COMPLETED_v1-v2.md))
+> **Statut :** 🔵 En cours — **P0 livré le 2026-09-27** (BUG-085 → BUG-088), **A5/A10/A12 livrés le 2026-09-28** (avec BUG-089), **A8/A9/A9bis livrés le 2026-09-28** (avec BUG-090), reste A6-A7 puis A13-A17
+> **Effort estimé :** 8-13 jours au total (P0 ✅ 2-3 j · P1 4-6 j · P2 2-4 j)
+> **Règle de maintenance :** la Roadmap porte les cases à cocher (suivi), cette fiche porte
+> l'analyse, les risques et les critères d'acceptation. **Ne pas dupliquer le détail.**
+
+---
+
+## 1. Périmètre et architecture
+
+| Couche | Fichier | Rôle |
+|---|---|---|
+| Lecture | `backend/xlsx_reader.py` | `render_sheets()` → un tableau HTML par feuille (openpyxl `read_only=True`, `data_only=False`) |
+| Endpoint lecture | `backend/routers/files_read.py:241-265` | `GET /api/file/{vault}?path=…` → `is_xlsx: true` + `xlsx_sheets: [{name, html, rows, cols, total_*, max_*, truncated}]` |
+| Endpoint fenêtre | `backend/routers/files_read.py` | `GET /api/file/{vault}/xlsx/sheet?path=&sheet=&offset=&limit=` (#153 A9) — une fenêtre de lignes, vraies coordonnées A1 |
+| Schéma API | `backend/schemas.py:286-290` | `is_xlsx`, `xlsx_sheets`, `XlsxSheetWindowResponse` |
+| Écriture | `backend/services/mutations.py:227-320` | `edit_xlsx_cells()` (backup, refs A1 validées, coercion `str`→`int`/`float`) |
+| Endpoint écriture | `backend/routers/files_write.py:116-148` | `PUT /api/file/{vault}/xlsx/save` (1 à 500 cellules / requête) |
+| Documentation API | `backend/openapi_docs.py:184-187` | exemple d'appel `xlsx/save` |
+| UI | `frontend/js/viewer.js:998-1100` | `renderXlsxViewer()` (onglets, cellules sales, Entrée/Échap, collage monoligne) |
+| CSS | `frontend/style.css:10927-10988` | `.xlsx-*` (variables CSS, colonne A `sticky`) |
+| Indexation | `backend/indexer.py:68, 563-568, 957-960` | `.xlsx` supporté, **métadonnées seules** (`content=""`) |
+| Outils IA | `backend/tools/documents.py:66-89` + `schemas.py:296-305` | `create_xlsx` (WRITE + confirmation) — **création seule** |
+| Tests | `tests/test_xlsx_viewer.py` | 11 tests backend (affichage, index, save, backup, 400) |
+
+## 2. Ce qui est supporté aujourd'hui (livré, non concerné par #153 sauf mention)
+
+**Lecture** — multi-feuilles avec onglets ; en-têtes A1/A2/B1 et numéros de ligne ; valeurs
+`_fmt()` (dates `YYYY-MM-DD` / `YYYY-MM-DD HH:MM`) ; lignes et colonnes de fin élaguées
+(`_trim`) ; feuille vide affichée ; `html.escape()` sur chaque valeur.
+
+**Édition** — `contentEditable` par `<td>`, classe `xlsx-dirty`, bouton Save actif seulement si
+modification ; `Entrée` → blur, `Échap` → restauration, collage forcé en monoligne ; un `PUT` par
+feuille sale ; coercion automatique des nombres (`"250"` → int `250`) ; chaîne vide → cellule
+vidée ; backup `.bak` avant écriture ; garde-fou vault read-only (403) ; `resolve_safe_path()`
+(anti path-traversal) ; `check_vault_access()` + `require_auth` ; journalisation d'audit
+(`log_file_save`).
+
+**Divers** — téléchargement de l'original ; refresh de l'arborescence via le watcher après
+écriture ; rafraîchissement de la visionneuse après une action IA (`create_xlsx` →
+`obsigate:file-written`, BUG-076).
+
+## 3. Limites connues (par couche)
+
+### 3.1 Fidélité du round-trip — risque n°1
+
+`load_workbook()` → `wb.save()` : ce qui est **réellement** perdu a été mesuré sur
+openpyxl 3.1.5 (2026-09-27), et non repris de la documentation :
+
+| Élément | Round-trip openpyxl 3.1.5 |
+|---|---|
+| Graphiques, images, dessins | ✅ **préservés** (mesuré : `xl/charts/`, `xl/drawings/`, `xl/media/` intacts) |
+| Tableaux croisés (pivot) + caches | ✅ **préservés** (`reader/excel.py` relit les `TableDefinition`, `workbook/_writer.py` les réécrit) |
+| Styles, formats, fusions, validation de données, mise en forme conditionnelle, commentaires | ✅ préservés |
+| **Valeurs calculées en cache** (`<f>…</f><v>…</v>`) | ❌ **perdues** → tout lecteur `data_only=True` (pandas, script tiers, convertisseur) renvoie `None` tant qu'Excel n'a pas recalculé |
+| Slicers / chronologies, contrôles de formulaire (`ctrlProps`/`activeX`), connexions & requêtes, custom XML, signature numérique, commentaires enrichis, macros | ❌ **perdus** (parties absentes de l'archive après écriture) |
+
+La liste fait foi dans le code : [`LOSSY_PARTS`](../backend/xlsx_reader.py) + la sonde
+`<f>…</f><v>[^<]` pour les valeurs en cache (openpyxl écrivant lui-même un `<v></v>` vide).
+
+**Ce qui reste ouvert** (non mesuré, prudence) : types de graphiques exotiques (treemap,
+sunburst, funnel…), `sparklines`, `xl/queryTables` en lecture Excel. Un classeur qui en contient
+peut sortir dégradé, voire échouer au chargement — d'où le refus par défaut (A1).
+
+### 3.2 Lecture
+
+- Aucun style, format de nombre, devise, pourcentage, largeur de colonne, ligne figée, cellule
+  fusionnée, commentaire, lien hypertexte, validation de données, mise en forme conditionnelle.
+- Plafonds durs `MAX_ROWS = 500`, `MAX_COLS = 40` par feuille, **sans indicateur dans l'UI** : au-delà,
+  contenu silencieusement tronqué et **non éditable**.
+- Pas de pagination ni de chargement à la demande : toutes les feuilles sont rendues d'un bloc
+  dans le JSON (20 feuilles × 20 000 cellules = payload énorme, UI gelée).
+- Formules affichées **en texte** (`=B1*2`), jamais recalculées ; après édition, les cellules
+  dépendantes ne se mettent pas à jour à l'écran.
+
+### 3.3 UI (`viewer.js`)
+
+Navigation clavier (Tab/flèches) absente ; pas de barre de formule, pas de nom de cellule actif,
+pas d'undo/redo global, pas de recherche dans la feuille, pas de tri/filtre, pas d'export CSV,
+pas d'ajout/renommage/suppression de feuille, pas d'insertion/suppression de ligne ou colonne,
+pas de sélection de plage, pas de copie d'une plage, pas de retour ligne dans une cellule
+(`Maj+Entrée`) ; seul le retour de l'API est signalé (plafond 500 cellules) ; seule la
+**colonne A** est `sticky` (le `thead` ne l'est pas → les en-têtes de colonnes disparaissent au
+défilement vertical). **Couverture de test** : `tests/frontend/xlsx-viewer.test.mjs` (10) et
+`tests/e2e/xlsx-viewer.spec.js` (3) depuis #153 P0 — la navigation clavier et la barre de formule
+restent à faire (A7).
+
+### 3.4 Recherche, IA et knowledge base
+
+- **Indexation** : `content=""` → un `.xlsx` est totalement **invisible** à la recherche TF-IDF, à
+  la recherche sémantique, au remplacement global, aux tags et aux statistiques de contenu.
+- **Outils IA** : seul `create_xlsx` existe (crée un fichier neuf, une seule feuille,
+  `overwrite=True` par défaut) ; `read_file` fait un `read_text()` sur l'archive ZIP → **bruit
+  binaire** envoyé au LLM ; pas de `update_xlsx_cells` pourtant le service existe déjà, pas
+  d'ajout de lignes, pas de `xlsx → markdown` pour le contexte.
+
+## 4. Risques de sécurité / robustesse
+
+| # | Risque | Où | Traitement | État |
+|---|---|---|---|---|
+| R1 | Perte silencieuse (valeurs calculées, slicers, contrôles, connexions, custom XML, signature) | `mutations.edit_xlsx_cells` | **A1** — bandeau + **409** `xlsx_lossy_content` sans `force` | 🟢 livré (BUG-085) |
+| R2 | Écriture non atomique (`wb.save()` en place) → classeur corrompu si crash | `mutations.edit_xlsx_cells` | **A2** — `.tmp` + `os.replace` | 🟢 livré (BUG-086) |
+| R3 | Concurrence : deux éditions (onglets, watcher + IA) → dernier écrivain gagne | `mutations.edit_xlsx_cells` | **A3** — verrou par chemin, **409** `conflict` | 🟢 livré (BUG-087) |
+| R4 | **Injection de formule** : une saisie `=cmd\|…`, `=HYPERLINK(…)` est stockée comme formule par openpyxl → DDE à l'ouverture dans Excel | `mutations._write_cell` | **A4** — forçage texte (`data_type="s"`), opt-in `allow_formula` | 🟢 livré (BUG-088) |
+| R5 | Troncature silencieuse au-delà de 500×40 | `xlsx_reader.MAX_ROWS/MAX_COLS` | A8 / A9 | 🟢 bandeau + dimensions exposées (BUG-090) ; le chargement paresseux par fenêtres sert les lignes au-delà du plafond |
+
+## 5. Backlog #153 — sous-tâches
+
+Légende : 🔴 P0 (sécurité / perte de données) · 🟡 P1 (valeur immédiate) · 🟢 P2 (confort /
+couverture) · effort en jours-homme de développement + tests.
+
+### P0 — Garde-fous d'écriture (2-3 j) — 🟢 livré le 2026-09-27
+
+- [x] **A1 — Alerte de fidélité avant écriture (R1).** `inspect_workbook()` liste ce qu'un
+  round-trip perd (`LOSSY_PARTS` + sonde valeurs en cache) ; la lecture renvoie
+  `xlsx_lossy_features` ; la visionneuse affiche un bandeau listant les éléments ; `PUT
+  …/xlsx/save` répond **409** `xlsx_lossy_content` (avec `details.features`) tant que `force` n'est
+  pas passé, le client demande confirmation puis réémet avec `force: true` (une seule fois par
+  session). *Vérifié :* `TestXlsxLossyGuard` (5), `xlsx-viewer.test.mjs` (10), E2E (3).
+- [x] **A2 — Écriture atomique (R2).** `wb.save(<nom>.<pid>.tmp)` puis `os.replace()` ; `.tmp`
+  supprimé sur échec ; backup `.bak` inchangé. Le `.tmp` est ignoré par le watcher. *Vérifié :*
+  `TestXlsxAtomicWrite` (2) — les octets d'origine sont intacts après un `save` en échec.
+- [x] **A3 — Verrou par fichier (R3).** Verrou `threading.Lock` par chemin (registre + garde,
+  timeout 15 s) autour du cycle load → edit → replace ; **409** `conflict` si le délai est dépassé.
+  L'endpoint est passé en `def` (sync) pour que l'attente s'exécute dans le threadpool. *Vérifié :*
+  `TestXlsxWriteLock` (2). *Limite :* verrou en mémoire, par processus (suffisant pour un serveur
+  ObsiGate, y compris desktop).
+- [x] **A4 — Neutralisation de l'injection de formule (R4).** `cell.data_type = "s"` après
+  affectation : une saisie `=`/`@` est stockée en texte. Opt-in `allow_formula: true` côté API et
+  bouton `f(x)` dans la visionneuse (état de session, jamais persisté). `+`/`-` restent des
+  nombres. Au passage : le handler `ServiceError` expose `code` + `details` et `api()` les
+  propage sur l'Error. *Vérifié :* `TestXlsxFormulaGuard` (4) + test du toggle côté UI.
+
+### P1 — Recherche, IA, UX (4-6 j) — 🟢 A5, A10, A12 livrés le 2026-09-28
+
+- [x] **A5 — Indexation du contenu des feuilles.** `extract_indexable_text()` (noms de feuilles +
+      20 premières lignes, `MAX_INDEX_CHARS = 5 000`, 20 feuilles max) alimente le TF-IDF et la
+      recherche sémantique ; la lecture binaire reste inchangée pour l'affichage. Un classeur
+      chiffré/corrompu s'indexe par son seul nom (jamais d'exception). Au passage : **BUG-089**,
+      un reindex manuel ne reconstruisait pas l'index inversé. *Vérifié :* `TestXlsxSearchable` (4)
+      + `TestXlsxIndexing`, **contre-preuve** (neutraliser l'extraction → 3 tests échouent).
+- [ ] **A6 — Outils IA sur classeur.** `update_xlsx_cells` (enveloppe du service existant),
+      `append_xlsx_rows`, `xlsx_to_markdown` (contexte LLM, plafonné), `list_xlsx_sheets` — risque
+      WRITE + confirmation pour les mutations, libellés i18n dans `backend/tools/labels.py`,
+      refresh viewer via `obsigate:file-written`.
+- [ ] **A7 — Navigation clavier & barre de formule.** `Tab`/`Maj+Tab`/`Entrée`/flèches, cellule
+      active affichée (nom A1), `Maj+Entrée` pour le multiligne, copier une plage, focus visible
+      et compatible mobile (≥ 44 px, `tests/e2e/mobile-editor.spec.js`).
+- [x] **A8 — `thead` sticky + indicateur de troncature (R5) — livré 2026-09-28 (BUG-090).**
+      Ligne d'en-têtes figlée au défilement vertical (`thead th { top: 0 }` ; `top: auto` sur les
+      numéros de ligne, sans quoi ils s'empilent en haut à gauche) ; `render_sheets()` expose
+      `total_rows`/`total_cols` (dimensions déclarées), `max_rows`/`max_cols` (plafonds) et
+      `truncated` — le bandeau « feuille tronquée » annonce le **plafond atteint** et non la
+      taille élaguée (une feuille creuse rend 1×1 tout en couvrant 500 lignes) ; libellés
+      `xlsx.truncated_*` FR/EN. *Vérifié :* `TestXlsxTruncationNotice` (4), `xlsx-viewer.test.mjs`
+      (4 nouveaux), E2E sur `test_vault/sample-xlsx-large.xlsx` (520 lignes).
+- [x] **A9 — Chargement paresseux par feuille (côté API).** Endpoint
+      `GET /api/file/{vault}/xlsx/sheet?sheet=&offset=&limit=` (`XlsxSheetWindowResponse`,
+      exemple dans `backend/openapi_docs.py`) : une fenêtre de 1 à 1 000 lignes (plafond
+      `MAX_WINDOW_ROWS`, `limit>1000` → 422), `has_more` pour paginer, valeurs calculées A12
+      incluses. Les numéros de ligne et `data-cell` restent les coordonnées A1 réelles de la
+      feuille (`_table(..., row_offset=offset)`) : une fenêtre est indistinguishable d'un rendu
+      complet et une édition dans la fenêtre cible la bonne cellule. Erreurs : 404 feuille
+      inconnue / fichier absent, 415 non-`.xlsx`. *Vérifié :* `TestXlsxSheetWindow` (11),
+      **contre-preuve** (neutraliser l'offset → 3 tests échouent), E2E « l'endpoint de fenêtre
+      sert les lignes au-delà du plafond ».
+- [x] **A9bis — Chargement à la demande côté UI.** Sous une feuille tronquée, un pied de page
+      « N lignes affichées sur M · Charger la suite » apparaît : cliquer — ou approcher du bas
+      du tableau (sentinelle de défilement, marge 120 px) — fetch la fenêtre suivante
+      (`limit=500`) et l'insère dans la table. Les lignes ajoutées passent par le **même**
+      pipeline d'édition que le rendu initial (`setupCell` factorisé : contenteditable, dirty,
+      Échap, collage monoligne, info-bulle valeurs calculées) et sont donc sauvegardables
+      immédiatement. Un fetch échoué restore le libellé du pied de page (retry possible) et
+      toast l'erreur ; feuille complète → pied de page masqué (`class="done"`).
+      *Vérifié :* `xlsx-viewer.test.mjs` 19/19 (5 nouveaux), **contre-preuve** (désactiver
+      `wireLazyRows` → 5 tests échouent), E2E « le bouton charger la suite ajoute les lignes
+      cachées » sur `sample-xlsx-large.xlsx` (A520 visible et éditable après clic).
+- [x] **A10 — Types et formats de saisie.** `_coerce_xlsx_value()` reconnait les booléens
+      (`true`/`vrai`/`oui`/`yes` et leurs négatifs) et les dates FR `JJ/MM/AAAA` (+ `HH:MM`),
+      jour-first comme Excel en locale française : `01/02/2026` = 1ᵉʳ février. Une saisie
+      ressemblant à une formule n'est jamais convertie (BUG-088 préservé) ; un code postal
+      numérique ou une version restent ce qu'ils sont. *Vérifié :* `TestXlsxValueCoercion` (5),
+      **contre-preuve** (neutraliser la coercion → 2 tests échouent).
+- [ ] **A11 — Tests frontend + E2E.** `tests/frontend/xlsx-viewer.test.mjs` (dirty, Échap,
+      collage, 1 PUT par feuille, bouton désactivé) et `tests/e2e/xlsx-viewer.spec.js`
+      (ouverture, onglets, édition, sauvegarde, rechargement) ; intégration au CI.
+- [x] **A12 — Valeurs calculées.** La valeur en cache s'affiche sous la formule dans un
+      `<span class="xlsx-cached">`. La 2ᵉ lecture `data_only=True` n'a lieu que si l'archive
+      contient réellement un `<f>…</f><v>…</v>` (sonde déjà présente pour A1) : le cas courant
+      reste à un seul chargement, et toute erreur retombe sur l'affichage formules seul.
+      L'info-bulle est traduite côté client (`xlsx.cached_value_title` FR/EN) — aucun texte
+      d'interface n'est émis par le backend. *Vérifié :* `TestXlsxCachedValues` (3),
+      **contre-preuve** (neutraliser la 2ᵉ lecture → 2 tests échouent).
+
+### P2 — Étendu (2-4 j)
+
+- [ ] **A13 — Tri / filtre / recherche dans la feuille + export CSV de la sélection.**
+- [ ] **A14 — CRUD de feuilles et de lignes/colonnes** (renommer, insérer, supprimer, dupliquer).
+- [ ] **A15 — Styles minimaux en écriture et lecture fidèle** (gras, fond, format
+      devise/pourcentage/date, cellules fusionnées, volets figés) ; conserver `csv-table` comme
+      socle de rendu.
+- [ ] **A16 — Formats additionnels.** `.xlsm` (`keep_vba=True`), `.xls`, `.ods`, `.csv` éditable
+      comme tableur — dépendances à qualifier (`xlrd`/`odfpy`) ou conversion.
+- [ ] **A17 — Vue « tableau de bord ».** Détection des plages nommées, TCD et graphiques ; vue
+      résumée (KPI par feuille) et proposal d'actions IA sur ces plages.
+
+## 6. Règles de livraison (rappel `AGENTS.md` / `DELIVERY_WORKFLOW.md`)
+
+- Chaque sous-tâche démarre par un **ID stable** : nouvelle feature = `#153-A<n>` dans la
+  Roadmap ; si la sous-tâche est un **défaut** (A1, A2, A3, A4, A8), l'ouvrir aussi comme
+  `BUG-NNN` dans `docs/ISSUES_TODOLIST.md` au moment du démarrage.
+- Backend : docstrings, `response_model` pour tout endpoint ajouté, exemple dans
+  `backend/openapi_docs.py`, chemin utilisateur via `resolve_safe_path()`.
+- Frontend : vanilla JS sans build, `safeCreateIcons()`, **variables CSS** (jamais de couleur
+  hardcodée), **i18n FR + EN** pour chaque nouveau texte (`test_i18n_parity.py` vert).
+- Tests : `pytest tests/test_xlsx_viewer.py`, `ruff`, `mypy`, `validate-imports`, suite frontend
+  ciblée, E2E si l'UI change — puis CI verte.
+- Documentation : `CHANGELOG.md` `[Unreleased]`, Roadmap (case cochée), cette fiche (résultat),
+  guide utilisateur i18n + README si impact utilisateur.
+
+## 7. Historique
+
+| Date | Événement |
+|---|---|
+| 2.27.0 | #152 livré : affichage multi-feuilles, édition des cellules, téléchargement (`docs/archive/COMPLETED_v1-v2.md`) |
+| 2026-09-27 | Audit complet → création de #153 : limites, risques R1-R5, backlog A1-A17 |
+| 2026-09-27 | Périmètre de perte **remesuré** sur openpyxl 3.1.5 : graphiques / images / TCD sont préservés, seules les valeurs en cache et quelques parties exotiques sont perdues |
+| 2026-09-27 | **P0 livré** (BUG-085 → BUG-088) : `xlsx_lossy_features` + 409 `xlsx_lossy_content`, écriture atomique, verrou par fichier, formules stockées en texte par défaut |
+| 2026-09-28 | **A5 + A10 + A12 livrés** : le contenu des cellules est indexé (recherche), la saisie est typée (booléens, dates FR), la valeur calculée s'affiche sous la formule. **BUG-089** corrigé au passage (reindex manuel ≠ reconstruction de l'index inversé ; `backend/search.py` lisait l'index par valeur) |
+| 2026-09-28 | **A8 + A9 livrés** (BUG-090) : la troncature d'une feuille est annoncée (bandeau + dimensions dans la réponse de lecture), les en-têtes restent visibles au défilement, et `GET …/xlsx/sheet` sert une fenêtre de lignes avec les vraies coordonnées A1 — les lignes au-delà du plafond redeviennent accessibles aux clients API. Défilement virtuel côté UI à suivre |
+| 2026-09-28 | **A9bis livré** : « Charger la suite » + sentinelle de défilement sous une feuille tronquée ; les lignes ajoutées sont éditables et sauvegardables immédiatement (même pipeline que le rendu initial) |
@@ -1551,6 +1551,17 @@
                    <nav class="help-nav" id="config-nav">
                        <div class="help-nav-header">
                            <div class="help-nav-title" data-i18n="config.title">Configuration</div>
+                            <button
+                                class="help-toc-close"
+                                id="config-toc-close"
+                                data-i18n-attr="aria-label:config.toc_close"
+                                aria-label="Fermer le sommaire"
+                            >
+                                <i
+                                    data-lucide="x"
+                                    style="width: 16px; height: 16px"
+                                ></i>
+                            </button>
                        </div>
                        <div class="help-nav-search-wrap">
                            <input
@@ -1565,17 +1576,6 @@
                                class="help-search-clear"
                                id="config-search-clear"
                                title="Effacer"
-                                onclick="
-                                    var s =
-                                        document.getElementById(
-                                            'config-nav-search',
-                                        );
-                                    if (s) {
-                                        s.value = '';
-                                        s.dispatchEvent(new Event('input'));
-                                        s.focus();
-                                    }
-                                "
                            >
                                X
                            </button>
@@ -1649,6 +1649,27 @@
                                                <span class="profile-hint" data-i18n="config.avatar_hint">PNG, JPG ou WEBP — image carrée recadrée et réduite à 256 px.</span>
                                            </div>
                                        </div>
+                                        <span class="profile-avatar-label" data-i18n="config.avatar_presets_label">Avatars prédéfinis</span>
+                                        <div
+                                            class="profile-avatar-presets"
+                                            id="profile-avatar-presets"
+                                            role="radiogroup"
+                                            data-i18n-attr="aria-label:config.avatar_presets_label"
+                                            aria-label="Avatars prédéfinis"
+                                        >
+                                            <button type="button" class="profile-avatar-preset" data-avatar="Chat.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/Chat.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="chien.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/chien.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="elephan.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/elephan.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="hibou.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/hibou.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="koala.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/koala.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="lapin.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/lapin.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="lion.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/lion.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="ours.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/ours.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="penda.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/penda.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="pingouin.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/pingouin.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="raton.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/raton.jpg" alt="" loading="lazy" /></button>
+                                            <button type="button" class="profile-avatar-preset" data-avatar="tigre.jpg" role="radio" aria-checked="false"><img src="/static/icons/avatar/tigre.jpg" alt="" loading="lazy" /></button>
+                                        </div>
                                        <p class="profile-avatar-error hidden" id="profile-avatar-error" role="alert"></p>
                                    </div>
                                    <div class="profile-field">
@@ -1666,7 +1687,7 @@
                                        <input type="text" id="profile-name" class="config-input" placeholder="Votre nom" maxlength="60">
                                    </div>
                                    <button class="config-save-btn" id="profile-save" data-i18n="config.save">Enregistrer</button>
-                                    <button class="config-save-btn" id="profile-logout" style="background:var(--danger-bg);color:var(--danger);border-color:var(--danger);margin-left:8px" data-i18n="config.logout" onclick="if(window.handleLogout)window.handleLogout();else{doLogoutFallback()}">Déconnexion</button>
+                                    <button class="config-save-btn" id="profile-logout" style="background:var(--danger-bg);color:var(--danger);border-color:var(--danger);margin-left:8px" data-i18n="config.logout">Déconnexion</button>
                                    <span class="profile-saved" id="profile-saved" style="display:none" data-i18n="config.saved">✓ Sauvegardé</span>
                                </div>
                            </section>
@@ -2146,7 +2167,6 @@
                                        >Nombre max de tokens élargis par préfixe (10-200)</span
                                    >
                                </div>
-                            </section>
                                <div class="config-actions-row">
                                    <button class="config-btn-save"
                                        id="cfg-save-backend" data-i18n="help.shortcut_save">
@@ -2509,7 +2529,7 @@

                            <!-- Partages publics -->
                            <section class="config-section help-section" id="cfg-partages-publics">
-                                <h2 id="cfg-partages-publics">📤 Partages publics</h2>
+                                <h2>📤 Partages publics</h2>
                                <p class="config-description">
                                    Liens de partage publics pour des documents
                                    (lecture seule, sans authentification).
@@ -2517,23 +2537,22 @@
                                <div id="shares-list"></div>
                            </section>

-                                                                                                                <!-- Plugins -->
-                                                                                                                <section
-                                                                                                                    class="config-section help-section"
-                                                                                                                    id="cfg-plugins"
-                                                                                                                >
-                                                                                                                    <h2 data-i18n="config.section_plugins">🧩 Plugins</h2>
-                                                                                                                    <p class="config-description" data-i18n="plugins.description">
-                                                                                                                        Extend ObsiGate with custom renderers, search filters, and editor actions.
-                                                                                                                    </p>
-                                                                                                                    <div id="plugins-settings-container"></div>
-                                                                                                                </section>
+                            <!-- Plugins -->
+                            <section
+                                class="config-section help-section"
+                                id="cfg-plugins"
+                            >
+                                <h2 data-i18n="config.section_plugins">🧩 Plugins</h2>
+                                <p class="config-description" data-i18n="plugins.description">
+                                    Extend ObsiGate with custom renderers, search filters, and editor actions.
+                                </p>
+                            </section>

-                                                                                                                <!-- À propos -->
-                                                        <section
-                                                            class="config-section help-section"
-                                                            id="cfg-about"
-                                                        >
+                            <!-- À propos -->
+                            <section
+                                class="config-section help-section"
+                                id="cfg-about"
+                            >
                                <h2 data-i18n="auto.a3319169">📦 À propos</h2>
                                <div
                                    id="config-about"
@@ -2975,14 +2994,6 @@
                        id="help-hamburger"
                        title="Sommaire"
                        aria-label="Afficher le sommaire"
-                        onclick="
-                            var n = document.getElementById('help-nav');
-                            if (n) {
-                                var d = n.style.display;
-                                n.style.display =
-                                    d === 'none' || d === '' ? 'flex' : 'none';
-                            }
-                        "
                    >
                        <i
                            data-lucide="menu"
@@ -3047,17 +3058,6 @@
                                id="help-search-clear"
                                title="Effacer la recherche"
                                aria-label="Effacer"
-                                onclick="
-                                    var s =
-                                        document.getElementById(
-                                            'help-nav-search',
-                                        );
-                                    if (s) {
-                                        s.value = '';
-                                        s.dispatchEvent(new Event('input'));
-                                        s.focus();
-                                    }
-                                "
                            >
                                ✕
                            </button>
@@ -4635,6 +4635,13 @@
                                    chercher) ; les actions de modification demandent une
                                    confirmation avec aperçu des changements.
                                </li>
+                                <li data-i18n="help.assistant_agent_run">
+                                    Les actions s'affichent dans le fil : l'assistant regroupe les
+                                    modifications en une seule approbation (« Tout approuver ») et
+                                    met à jour l'arborescence et le document ouvert dès qu'elles
+                                    sont appliquées. Le bouton d'envoi devient « Stop » pour
+                                    interrompre l'exécution à tout moment.
+                                </li>
                                <li data-i18n="help.assistant_resize">
                                    Le bord gauche du panneau est redimensionnable ; la largeur est
                                    mémorisée.
@@ -5614,7 +5621,7 @@ curl -X POST https://votre-serveur.com/webhook \
                    data-lucide="folder-open"
                    style="width: 20px; height: 20px"
                ></i>
-                <span class="mt-label">settings.search</span>
+                <span class="mt-label" data-i18n="settings.explorer">Explorateur</span>
            </button>
            <button
                class="mt-btn"
@@ -72,14 +72,25 @@ async function api(path, opts) {
  }
  if (!res.ok) {
    var detail = "";
+    var code = "";
+    var details = null;
    try {
      var body = await res.json();
      detail = body.detail || "";
+      // #153 A1 : the service layer exposes a stable code + details so callers
+      // can branch on the failure (e.g. confirm a lossy .xlsx write) instead of
+      // matching on the message.
+      code = body.code || "";
+      details = body.details || null;
    } catch (_) {
      /* no json body */
    }
    showToast(detail || "Erreur API : " + res.status, "error");
-    throw new Error(detail || "API error: " + res.status);
+    var apiError = new Error(detail || "API error: " + res.status);
+    apiError.status = res.status;
+    apiError.code = code;
+    apiError.details = details;
+    throw apiError;
  }
  return res.json();
 }
@@ -1305,8 +1316,7 @@ async function _startMfaSetup() {
    // secret when the backend has no QR generator available.
    const qrImg = data.qr_data_url
      ? `<img id="mfa-qr-img" alt="QR Code" class="mfa-qr-code-img"
-              src="${data.qr_data_url}"
-              onerror="this.style.display='none';document.getElementById('mfa-qr-fallback').style.display='block';">`
+              src="${data.qr_data_url}">`
      : "";
    const fallbackStyle = data.qr_data_url ? "display:none" : "";
    flowArea.innerHTML = `
@@ -1335,6 +1345,17 @@ async function _startMfaSetup() {
      codeInput.value = codeInput.value.replace(/[^0-9]/g, "");
    });

+    // QR fallback (#87, ex-onerror inline) : si l'image ne charge pas,
+    // afficher la saisie manuelle du secret.
+    const qrImgEl = document.getElementById("mfa-qr-img");
+    if (qrImgEl) {
+      qrImgEl.addEventListener("error", () => {
+        qrImgEl.style.display = "none";
+        const fallback = document.getElementById("mfa-qr-fallback");
+        if (fallback) fallback.style.display = "block";
+      });
+    }
+
    document.getElementById("mfa-confirm-btn").addEventListener("click", async () => {
      const code = codeInput.value.trim();
      if (code.length !== 6) return;
@@ -52,6 +52,24 @@ const PANEL_MIN_WIDTH = 320;
 const PANEL_MAX_WIDTH = 1000;
 const PANEL_WIDTH_KEY = 'obsigate-bookslm-width';

+// BUG-076 — Tools that mutate the vault: their execution must refresh the
+// sidebar tree immediately (the watcher SSE is delayed and directory-only
+// changes may not produce an index event).
+const MUTATING_TOOLS = new Set([
+  'create_file', 'create_directory', 'edit_file', 'append_to_file',
+  'rename_file', 'rename_directory', 'move_path', 'replace_in_files',
+  'delete_file', 'delete_directory', 'restore_backup',
+  'create_xlsx', 'create_docx', 'create_csv', 'create_pdf',
+  'update_xlsx_cells', 'append_xlsx_rows',
+]);
+// Subset carrying a concrete `vault` + `path`: the displayed document is
+// reloaded from disk so an open viewer/editor reflects the agent's write.
+const FILE_WRITE_TOOLS = new Set([
+  'edit_file', 'append_to_file', 'create_file', 'restore_backup',
+  'create_xlsx', 'create_docx', 'create_csv', 'create_pdf',
+  'update_xlsx_cells', 'append_xlsx_rows',
+]);
+
 /**
 * BUG-059 — Is this pointer press a scrollbar drag (and only that)?
 *
@@ -895,15 +913,14 @@ class BooksLM {
  }

  /**
-   * #93 — A write tool just modified a vault file. Notify the UI so the
-   * displayed document is reloaded from disk: otherwise the read view keeps the
-   * stale content, and an open editor buffer autosaves the old text back over
-   * the assistant's change (see utils.reloadExternalWrite).
+   * #93 / BUG-076 — A write tool just modified a vault file. Notify the UI so
+   * the displayed document is reloaded from disk: otherwise the read view keeps
+   * the stale content, and an open editor buffer autosaves the old text back
+   * over the assistant's change (see utils.reloadExternalWrite).
   */
  _notifyFileWritten(data) {
    if (!data || data.ok === false) return;
-    const WRITE_TOOLS = ['edit_file', 'append_to_file', 'create_file', 'restore_backup'];
-    if (WRITE_TOOLS.indexOf(data.name) === -1) return;
+    if (!FILE_WRITE_TOOLS.has(data.name)) return;
    const args = data.arguments || {};
    if (!args.vault || !args.path) return;
    window.dispatchEvent(
@@ -913,6 +930,24 @@ class BooksLM {
    );
  }

+  /**
+   * BUG-076 — Refresh the sidebar tree right after an agent mutation, without
+   * waiting for the (debounced) watcher SSE. Debounced so a batch of tool
+   * events (e.g. a folder plus its files) triggers a single refresh.
+   */
+  _scheduleTreeRefresh() {
+    if (this._treeRefreshTimer) clearTimeout(this._treeRefreshTimer);
+    this._treeRefreshTimer = setTimeout(async () => {
+      this._treeRefreshTimer = null;
+      try {
+        const m = await import('./sidebar.js');
+        if (m && typeof m.refreshSidebarTreePreservingState === 'function') {
+          await m.refreshSidebarTreePreservingState();
+        }
+      } catch { /* tree refresh is best-effort */ }
+    }, 250);
+  }
+
  // ── Rendering ───────────────────────────────────────────────────────

  _render() {
@@ -1041,7 +1076,11 @@ class BooksLM {
      }
    });

-    panel.querySelector('.bookslm-btn-send').addEventListener('click', () => this._sendMessage());
+    panel.querySelector('.bookslm-btn-send').addEventListener('click', () => {
+      // BUG-077: the same button stops the run while a response streams.
+      if (this._isLoading) this._stopGeneration();
+      else this._sendMessage();
+    });

    // Resizable panel (drag the left edge).
    const resizeHandle = panel.querySelector('.bookslm-resize-handle');
@@ -2665,6 +2704,17 @@ class BooksLM {
      return t('ai.tool_call', { name: call.name });
    }

+  /**
+   * BUG-074 — number of *action* steps of a message (reasoning notes are not
+   * actions). Falls back to the total when the block holds only thoughts so a
+   * “0 étape” label can never appear.
+   */
+  _actionStepCount(toolCalls) {
+    const list = toolCalls || [];
+    const actions = list.filter((c) => !(c.step && c.step.key === 'thought'));
+    return actions.length || list.length;
+  }
+
  /** Chevron used by every collapsible block (▶ closed / ▼ open, via CSS). */
  _chevron() {
    const c = document.createElement('span');
@@ -2698,13 +2748,32 @@ class BooksLM {
      dots.classList.add('bookslm-steps-dots');
      summary.appendChild(dots);
    }
-    const countKey = toolCalls.length > 1 ? 'ai.steps_count_plural' : 'ai.steps_count';
+    // BUG-074: the counter reflects *actions*, not reasoning notes, and the
+    // collapsed header also previews the first action so an “N étapes ▶” line
+    // is no longer an opaque title.
+    const actionCalls = toolCalls.filter((c) => !(c.step && c.step.key === 'thought'));
+    const count = this._actionStepCount(toolCalls);
+    const countKey = count > 1 ? 'ai.steps_count_plural' : 'ai.steps_count';
    const label = document.createElement('span');
    label.className = 'bookslm-steps-label';
-    label.textContent = t(countKey, { count: toolCalls.length });
+    label.textContent = t(countKey, { count });
    summary.appendChild(label);
+    const first = actionCalls[0];
+    let preview = '';
+    if (first) {
+      preview = this._stepText(first);
+      const title = document.createElement('span');
+      title.className = 'bookslm-steps-title';
+      title.textContent = `— ${preview}`;
+      summary.appendChild(title);
+    }
    summary.appendChild(this._chevron());
-    if (running) summary.setAttribute('aria-label', `${label.textContent} — ${t('ai.steps_running')}`);
+    if (running) {
+      summary.setAttribute(
+        'aria-label',
+        `${label.textContent}${preview ? ` — ${preview}` : ''} — ${t('ai.steps_running')}`,
+      );
+    }
    wrap.appendChild(summary);

    const body = document.createElement('div');
@@ -2804,8 +2873,11 @@ class BooksLM {
    const conf = msg.confirmation;
    const pending = conf.pending || {};
    const error = pending.error || pending;
-    const tool = error.tool || 'action';
-    const args = error.arguments || {};
+    // BUG-075: the run batches every mutating call of the LLM turn into
+    // `pending.actions`; fall back to the single legacy `error` shape.
+    const actions = (Array.isArray(pending.actions) && pending.actions.length)
+      ? pending.actions
+      : [{ id: error.id, tool: error.tool, arguments: error.arguments || {}, step: null }];

    const card = document.createElement('div');
    card.className = 'bookslm-action bookslm-confirm';
@@ -2814,23 +2886,48 @@ class BooksLM {
    meta.className = 'bookslm-action-meta';
    meta.innerHTML = '<span class="bookslm-action-icon">🔒</span>';
    const textEl = document.createElement('span');
-    textEl.textContent = t('ai.tool_call', { name: tool }) + (args.path ? ` — ${args.path}` : '');
+    if (actions.length > 1) {
+      textEl.textContent = t('ai.confirm_actions', { count: actions.length });
+    } else {
+      const tool = actions[0].tool || 'action';
+      const args = actions[0].arguments || {};
+      textEl.textContent = t('ai.tool_call', { name: tool }) + (args.path ? ` — ${args.path}` : '');
+    }
    meta.appendChild(textEl);
    card.appendChild(meta);

-    const diffHost = document.createElement('div');
-    diffHost.className = 'bookslm-confirm-diff';
-    if (conf._diffHtml) {
-      diffHost.innerHTML = conf._diffHtml;
-    } else if (!conf._diffLoading) {
-      conf._diffLoading = true;
-      this._fillConfirmationDiff(args, conf).then(() => this._renderMessages());
+    // One row per action with a diff preview when it writes file content.
+    const hasContent = (a) => {
+      const args = a.arguments || {};
+      return typeof args.content === 'string' && args.vault && args.path;
+    };
+    if (actions.length > 1) {
+      const list = document.createElement('div');
+      list.className = 'bookslm-confirm-actions';
+      for (const action of actions) {
+        const args = action.arguments || {};
+        const row = document.createElement('details');
+        row.className = 'bookslm-confirm-action';
+        const summary = document.createElement('summary');
+        const text = document.createElement('span');
+        text.textContent = this._stepText({ step: action.step, name: action.tool })
+          + (args.path ? ` — ${args.path}` : '');
+        summary.appendChild(text);
+        if (hasContent(action)) summary.appendChild(this._chevron());
+        row.appendChild(summary);
+        if (hasContent(action)) row.appendChild(this._diffHost(action, args));
+        list.appendChild(row);
+      }
+      card.appendChild(list);
+    } else if (hasContent(actions[0])) {
+      card.appendChild(this._diffHost(conf, actions[0].arguments || {}));
    }
-    card.appendChild(diffHost);

    const apply = document.createElement('button');
    apply.className = 'bookslm-action-apply';
-    apply.textContent = t('ai.action_apply');
+    apply.textContent = actions.length > 1
+      ? t('ai.action_apply_all', { count: actions.length })
+      : t('ai.action_apply');
    apply.addEventListener('click', async () => {
      apply.disabled = true;
      apply.textContent = t('ai.action_applying');
@@ -2839,7 +2936,9 @@ class BooksLM {
        apply.textContent = t('ai.action_applied');
      } catch (e) {
        apply.disabled = false;
-        apply.textContent = t('ai.action_apply');
+        apply.textContent = actions.length > 1
+          ? t('ai.action_apply_all', { count: actions.length })
+          : t('ai.action_apply');
        showToast(t('ai.action_failed', { error: e.message }), 'error');
      }
    });
@@ -2847,6 +2946,19 @@ class BooksLM {
    return card;
  }

+  /** Diff host for one confirmation action (lazy LCS diff, cached on the host). */
+  _diffHost(host, args) {
+    const diffHost = document.createElement('div');
+    diffHost.className = 'bookslm-confirm-diff';
+    if (host._diffHtml) {
+      diffHost.innerHTML = host._diffHtml;
+    } else if (!host._diffLoading) {
+      host._diffLoading = true;
+      this._fillConfirmationDiff(args, host).then(() => this._renderMessages());
+    }
+    return diffHost;
+  }
+
  async _fillConfirmationDiff(args, conf) {
    const proposed = args.content;
    if (typeof proposed !== 'string' || !args.vault || !args.path) return;
@@ -2935,22 +3047,20 @@ class BooksLM {
    if (!conf || conf._applying) return;
    conf._applying = true;

+    // BUG-075: one click applies the whole pending batch AND authorizes the
+    // remaining actions of the same run (no second confirmation card).
    const payload = {
      ...(msg.payload || {}),
      confirm: conf.pending,
      confirm_messages: conf.messages,
+      confirm_all: true,
    };

    this._isLoading = true;
    this._abortCtrl = new AbortController();
-    const sendBtn = this._panel && this._panel.querySelector('.bookslm-btn-send');
-    if (sendBtn) sendBtn.disabled = true;
+    this._syncSendButton();
+    this._setActivity('working', t('ai.activity_streaming'));

-    // BUG-046: carry the original request payload over to the continuation.
-    // When an applied tool failed and the model re-proposed a confirmation,
-    // the continuation used to have `payload: null`: the second "Appliquer"
-    // then resumed without `message` → 422 → "[object Object]" toast.
-    const continuation = { role: 'assistant', content: '', sources: [], toolCalls: [], confirmation: null, payload: msg.payload ? { ...msg.payload } : null };
    try {
      let resp = await this._postChat(payload);
      if (resp.status === 401 && AuthManager._authEnabled) {
@@ -2960,21 +3070,65 @@ class BooksLM {
      if (!resp.ok) {
        throw await this._responseError(resp);
      }
-      // The confirmation is resolved: drop the card and show the continuation.
+      // The confirmation is resolved: drop the card and keep streaming into the
+      // SAME message, so the steps block accumulates the whole exchange
+      // instead of fragmenting into a new “1 étape” message per approval
+      // (BUG-074). BUG-046 payload carry-over is inherent: `msg.payload` stays.
      msg.confirmation = null;
-      this._messages.push(continuation);
      this._renderMessages({ anchor: true });
-      await this._streamResponse(resp, continuation, payload);
+      await this._streamResponse(resp, msg, payload);
+    } catch (e) {
+      if (e.name === 'AbortError') {
+        this._markStopped(msg);
+      } else {
+        throw e;
+      }
    } finally {
      conf._applying = false;
      this._isLoading = false;
      this._abortCtrl = null;
-      if (sendBtn) sendBtn.disabled = false;
+      this._syncSendButton();
      this._renderMessages();
      this._saveHistory();
    }
  }

+  /** BUG-077 — abort the in-flight agent/chat request. */
+  _stopGeneration() {
+    if (this._abortCtrl) {
+      try {
+        this._abortCtrl.abort();
+      } catch { /* already aborted */ }
+    }
+  }
+
+  /** Append a visible « stopped by the user » marker to an assistant message. */
+  _markStopped(msg) {
+    const marker = `⏹ ${t('ai.stopped')}`;
+    const content = String(msg.content || '');
+    if (!content.includes(marker)) {
+      msg.content = content ? `${content}\n\n${marker}` : marker;
+    }
+    this._setActivity('idle');
+  }
+
+  /**
+   * BUG-077 — the composer's round button doubles as a Stop control while a
+   * response streams: a new send is already blocked by `_isLoading`, so the
+   * button switches to a stop icon instead of being disabled.
+   */
+  _syncSendButton() {
+    const btn = this._panel && this._panel.querySelector('.bookslm-btn-send');
+    if (!btn) return;
+    const loading = !!this._isLoading;
+    btn.classList.toggle('is-stopping', loading);
+    btn.title = loading ? t('ai.stop') : t('bookslm.send');
+    btn.setAttribute('aria-label', btn.title);
+    const icon = btn.querySelector('i');
+    if (icon) icon.setAttribute('data-lucide', loading ? 'square' : 'arrow-up');
+    if (typeof safeCreateIcons === 'function') safeCreateIcons();
+  }
+
  // ── Messaging ───────────────────────────────────────────────────────

  async _sendMessage() {
@@ -3053,10 +3207,8 @@ class BooksLM {
    this._isLoading = true;
    this._renderMessages({ anchor: true, instant: true });

-    const sendBtn = this._panel.querySelector('.bookslm-btn-send');
-    if (sendBtn) sendBtn.disabled = true;
-
    this._abortCtrl = new AbortController();
+    this._syncSendButton();
    this._setActivity('working', t('ai.activity_sending'));

    try {
@@ -3085,17 +3237,17 @@ class BooksLM {
      }
      this._setActivity('done', t('ai.activity_done'));
    } catch (e) {
-      if (e.name !== 'AbortError') {
+      if (e.name === 'AbortError') {
+        this._markStopped(assistantMsg);
+      } else {
        assistantMsg.content = `⚠ Error: ${e.message}`;
        console.warn('AI assistant chat error:', e);
        this._setActivity('error', t('ai.activity_error'));
-      } else {
-        this._setActivity('idle');
      }
    }

    this._isLoading = false;
-    if (sendBtn) sendBtn.disabled = false;
+    this._syncSendButton();
    this._abortCtrl = null;
    this._renderMessages();
    this._saveHistory();
@@ -3155,6 +3307,10 @@ class BooksLM {
        });
        // #93 — a vault write must be reflected in the displayed document.
        this._notifyFileWritten(data);
+        // BUG-076 — reflect tree changes (create/delete/rename…) right away.
+        if (data.ok !== false && MUTATING_TOOLS.has(data.name)) {
+          this._scheduleTreeRefresh();
+        }
        this._setActivity('working', t('ai.activity_tool', { name: data.name }));
        return;
      }
@@ -363,6 +363,27 @@ function initHelpModal() {
    }
  });

+  // Help TOC hamburger + search clear (#87, ex-onclick inline in index.html).
+  var helpHamburger = document.getElementById("help-hamburger");
+  if (helpHamburger) {
+    helpHamburger.addEventListener("click", function() {
+      var n = document.getElementById("help-nav");
+      if (n) {
+        var d = n.style.display;
+        n.style.display = d === "none" || d === "" ? "flex" : "none";
+      }
+    });
+  }
+  var helpSearch = document.getElementById("help-nav-search");
+  var helpSearchClear = document.getElementById("help-search-clear");
+  if (helpSearchClear && helpSearch) {
+    helpSearchClear.addEventListener("click", function() {
+      helpSearch.value = "";
+      helpSearch.dispatchEvent(new Event("input"));
+      helpSearch.focus();
+    });
+  }
+
  document.addEventListener("keydown", (e) => {
    if (e.key === "Escape" && modal.classList.contains("active")) {
      closeHelpModal();
@@ -767,10 +788,15 @@ function initConfigModal() {
  openBtn.addEventListener("click", async () => {
    modal.classList.add("active");
    closeHeaderMenu();
-    // BUG-071: reset the TOC to the CSS default (mobile: hidden, desktop:
-    // visible) like the help modal does on open.
+    // BUG-071/#114: reset the TOC to the CSS default (mobile: hidden drawer,
+    // desktop: visible sidebar) like the help modal does on open. Clear the
+    // stale inline display and the drawer-open class so a previous mobile
+    // session cannot leave the nav stuck open.
    var configNavOnOpen = document.getElementById("config-nav");
    if (configNavOnOpen) configNavOnOpen.style.display = '';
+    modal.classList.remove("config-toc-open");
+    var configHamburgerOnOpen = document.getElementById("config-hamburger");
+    if (configHamburgerOnOpen) configHamburgerOnOpen.classList.remove("active");
    renderConfigFilters();
    loadConfigFields();
    loadDiagnostics();
@@ -786,6 +812,12 @@ function initConfigModal() {
  closeBtn.addEventListener("click", closeConfigModal);
  modal.addEventListener("click", (e) => {
    if (e.target === modal) {
+      // #114: a tap on the backdrop (or outside the drawer) first closes the
+      // TOC drawer; only a second tap closes the whole modal.
+      if (modal.classList.contains("config-toc-open")) {
+        _setConfigNav(false);
+        return;
+      }
      closeConfigModal();
    }
  });
@@ -872,7 +904,7 @@ function initConfigModal() {
    });
  }

-  // Logout button — handled inline in index.html (onclick)
+  // Logout button — wired here with addEventListener (#87, no inline onclick)

  // Config nav search
  var cfgSearch = document.getElementById("config-nav-search");
@@ -890,17 +922,29 @@ function initConfigModal() {
    });
  }

-  // BUG-071: mobile table of contents. #config-nav shares the .help-nav
+  // Config search clear button (#87, ex-onclick inline).
+  var cfgSearchClear = document.getElementById("config-search-clear");
+  if (cfgSearchClear && cfgSearch) {
+    cfgSearchClear.addEventListener("click", function() {
+      cfgSearch.value = "";
+      cfgSearch.dispatchEvent(new Event("input"));
+      cfgSearch.focus();
+    });
+  }
+
+  // BUG-071/#114: mobile table of contents. #config-nav shares the .help-nav
  // rule that hides it below 768px, but — unlike the help modal — the config
  // modal had no toggle to reveal it, leaving mobile users with no way to
-  // reach a section. The header hamburger opens it as a top block; picking
-  // a section smooth-scrolls inside the modal and collapses it on mobile.
+  // reach a section. The header hamburger opens it as a left slide-over
+  // drawer (backdrop via .config-toc-open on the modal); picking a section
+  // smooth-scrolls inside the modal and collapses it on mobile.
  var configNav = document.getElementById("config-nav");
  var configHamburger = document.getElementById("config-hamburger");
  function _isConfigMobile() { return window.innerWidth <= 768; }
  function _setConfigNav(open) {
    if (!configNav) return;
    configNav.style.display = open ? "flex" : "none";
+    modal.classList.toggle("config-toc-open", !!open && _isConfigMobile());
    if (configHamburger) configHamburger.classList.toggle("active", !!open);
  }
  if (configHamburger) {
@@ -910,6 +954,14 @@ function initConfigModal() {
      _setConfigNav(hidden);
    });
  }
+  // #114: close button inside the drawer header.
+  var configTocClose = document.getElementById("config-toc-close");
+  if (configTocClose) {
+    configTocClose.addEventListener("click", function(e) {
+      e.stopPropagation();
+      _setConfigNav(false);
+    });
+  }
  if (configNav) {
    configNav.querySelectorAll(".help-nav-link").forEach(function(a) {
      a.addEventListener("click", function(e) {
@@ -930,6 +982,11 @@ function initConfigModal() {

  document.addEventListener("keydown", (e) => {
    if (e.key === "Escape" && modal.classList.contains("active")) {
+      // #114: Escape closes the TOC drawer first, then the modal.
+      if (modal.classList.contains("config-toc-open")) {
+        _setConfigNav(false);
+        return;
+      }
      closeConfigModal();
    }
  });
@@ -949,7 +1006,13 @@ function initConfigModal() {

 function closeConfigModal() {
  const modal = document.getElementById("config-modal");
-  if (modal) modal.classList.remove("active");
+  if (modal) {
+    modal.classList.remove("active");
+    // #114: never leave the drawer-open class (backdrop) behind.
+    modal.classList.remove("config-toc-open");
+    const nav = document.getElementById("config-nav");
+    if (nav) nav.style.display = '';
+  }
 }

 // --- Config field helpers ---
@@ -1159,7 +1222,7 @@ function renderDiagnostics(container, data) {
        ["Postings total", data.inverted_index.total_postings.toLocaleString()],
        ["Documents", data.inverted_index.documents],
        ["Mémoire estimée", data.inverted_index.memory_estimate_mb + " MB"],
-        ["Stale", data.inverted_index.is_stale ? "Oui" : "Non"],
+        ["Index prêt", data.inverted_index.is_ready ? "Oui" : "Non"],
      ],
    },
    {
@@ -1543,13 +1606,15 @@ export async function openShareDialog(vault, path) {
          <p style="font-size:0.85rem;color:var(--text-muted);margin-bottom:4px">${escapeHtml(vault)}/${escapeHtml(path)}</p>
          ${expiresInfo}
          <p style="font-size:0.75rem;color:var(--text-muted);margin-bottom:8px">${existingShare.access_count} vue(s)</p>
-          <input type="text" class="share-url-input" value="${url}" readonly onclick="this.select()">
+          <input type="text" class="share-url-input" value="${url}" readonly>
          <div class="share-dialog-actions">
            <button class="share-copy-btn">📋 Copier le lien</button>
            <button class="share-revoke-btn">🗑 Révoquer</button>
            <button class="share-close-btn">Fermer</button>
          </div>
        </div>`;
+      const shareUrlInput = div.querySelector(".share-url-input");
+      if (shareUrlInput) shareUrlInput.addEventListener("click", function() { shareUrlInput.select(); });
      div.querySelector(".share-copy-btn").addEventListener("click", async () => {
        try {
          await navigator.clipboard.writeText(url);
@@ -2493,7 +2558,10 @@ function initProfile() {
    } catch(e) {}
  });

-  // Logout button — handled inline in index.html (onclick)
+  // Logout button (#87, ex-onclick inline in index.html).
+  if (logoutBtn && window.handleLogout) {
+    logoutBtn.addEventListener('click', function() { window.handleLogout(); });
+  }

  // ── Avatar (#113) ────────────────────────────────────────────────
  var avatarField = document.getElementById('profile-avatar-field');
@@ -2502,6 +2570,7 @@ function initProfile() {
  var overlayBtn = document.getElementById('profile-avatar-overlay');
  var previewBox = document.getElementById('profile-avatar-preview');
  var removeBtn = document.getElementById('profile-avatar-remove');
+  var presetsBox = document.getElementById('profile-avatar-presets');
  if (!avatarField || !avatarInput || !chooseBtn) return;

  // The avatar only exists on a real account — hide it when auth is off.
@@ -2522,6 +2591,59 @@ function initProfile() {
    })
    .catch(function () { /* non-bloquant */ });

+  // ── Preset avatars (#117) ────────────────────────────────────────
+  var PRESET_STORAGE_KEY = 'obsigate-avatar-preset';
+  function markActivePreset(name) {
+    if (!presetsBox) return;
+    presetsBox.querySelectorAll('.profile-avatar-preset').forEach(function (btn) {
+      var on = !!name && btn.getAttribute('data-avatar') === name;
+      btn.classList.toggle('active', on);
+      btn.setAttribute('aria-checked', on ? 'true' : 'false');
+    });
+  }
+  function clearActivePreset() {
+    try { localStorage.removeItem(PRESET_STORAGE_KEY); } catch (e) { /* ignore */ }
+    markActivePreset(null);
+  }
+  function rememberActivePreset(name) {
+    try { localStorage.setItem(PRESET_STORAGE_KEY, name); } catch (e) { /* ignore */ }
+    markActivePreset(name);
+  }
+  var savedPreset = null;
+  try { savedPreset = localStorage.getItem(PRESET_STORAGE_KEY); } catch (e) { /* ignore */ }
+  markActivePreset(savedPreset);
+
+  if (presetsBox) {
+    presetsBox.querySelectorAll('.profile-avatar-preset').forEach(function (btn) {
+      btn.addEventListener('click', async function () {
+        var preset = btn.getAttribute('data-avatar');
+        if (!preset) return;
+        _profileAvatarError(null);
+        var all = presetsBox.querySelectorAll('.profile-avatar-preset');
+        all.forEach(function (b) { b.disabled = true; });
+        try {
+          // Load the bundled image, then reuse the upload pipeline (center-crop
+          // + 256 px JPEG) so it stores exactly like an imported picture.
+          var res = await fetch('/static/icons/avatar/' + encodeURIComponent(preset), { credentials: 'include' });
+          if (!res.ok) throw new Error('HTTP ' + res.status);
+          var blob = await res.blob();
+          var dataUrl = await _resizeAvatarFile(blob);
+          var user = await _patchProfileAvatar(dataUrl);
+          AuthManager.updateCachedUser(user);
+          _renderProfileAvatar(user.avatar || dataUrl);
+          AuthManager.renderUserSection();
+          rememberActivePreset(preset);
+          showToast(t('config.avatar_updated'), 'success');
+        } catch (err) {
+          _profileAvatarError('config.avatar_upload_failed');
+          console.error('Avatar preset failed:', err);
+        } finally {
+          all.forEach(function (b) { b.disabled = false; });
+        }
+      });
+    });
+  }
+
  function openPicker() {
    _profileAvatarError(null);
    avatarInput.click();
@@ -2555,6 +2677,7 @@ function initProfile() {
      AuthManager.updateCachedUser(user);
      _renderProfileAvatar(user.avatar || dataUrl);
      AuthManager.renderUserSection();
+      clearActivePreset();               // an imported photo is not a preset
      showToast(t('config.avatar_updated'), 'success');
    } catch (err) {
      _profileAvatarError('config.avatar_upload_failed');
@@ -2573,6 +2696,7 @@ function initProfile() {
        AuthManager.updateCachedUser(user);
        _renderProfileAvatar(null);
        AuthManager.renderUserSection();
+        clearActivePreset();
        showToast(t('config.avatar_removed'), 'success');
      } catch (err) {
        _profileAvatarError('config.avatar_upload_failed');
@@ -35,7 +35,7 @@ export function renderExcalidraw(container, data, vaultName, filePath, opts = {}
  // Build the iframe
  const iframe = document.createElement('iframe');
  iframe.id = editorId;
-  iframe.src = '/static/excalidraw-editor.html?v=' + Date.now();
+  iframe.src = '/excalidraw-editor.html?v=' + Date.now();
  iframe.sandbox.add('allow-scripts');
  iframe.sandbox.add('allow-same-origin');
  // Let the editor's own Fullscreen button work (native Fullscreen API inside
@@ -478,8 +478,8 @@ function openTemplateModal() {
                    '</div>' +
                '</div>' +
                '<div class="modal-footer">' +
-                    '<button class="btn btn-primary" data-i18n="common.close" onclick="this.closest(\'.modal-overlay\').remove()">Close</button>' +
-                    '<button class="btn btn-secondary" onclick="copyTemplate()">Copy to Clipboard</button>' +
+                    '<button class="btn btn-primary btn-modal-close" data-i18n="common.close">Close</button>' +
+                    '<button class="btn btn-secondary btn-copy-template">Copy to Clipboard</button>' +
                '</div>' +
            '</div>';

@@ -487,11 +487,11 @@ function openTemplateModal() {

        modal.querySelector('.modal-close')?.addEventListener('click', () => modal.remove());
        modal.addEventListener('click', (e) => { if (e.target === modal) modal.remove(); });
-
-        window.copyTemplate = () => {
+        modal.querySelector('.btn-modal-close')?.addEventListener('click', () => modal.remove());
+        modal.querySelector('.btn-copy-template')?.addEventListener('click', () => {
            navigator.clipboard.writeText(JSON.stringify(template, null, 2));
            showToast('Template copied to clipboard', 'success');
-        };
+        });
    });
 }

@@ -508,12 +508,13 @@ function openCodeModal(name, code) {
                '<pre class="code-block" style="max-height: 500px; overflow: auto;">' + escapeHtml(code) + '</pre>' +
            '</div>' +
            '<div class="modal-footer">' +
-                '<button class="btn btn-primary" data-i18n="common.close" onclick="this.closest(\'.modal-overlay\').remove()">Close</button>' +
+                '<button class="btn btn-primary btn-modal-close" data-i18n="common.close">Close</button>' +
            '</div>' +
        '</div>';

    document.body.appendChild(modal);
    modal.querySelector('.modal-close')?.addEventListener('click', () => modal.remove());
+    modal.querySelector('.btn-modal-close')?.addEventListener('click', () => modal.remove());
    modal.addEventListener('click', (e) => { if (e.target === modal) modal.remove(); });
 }

@@ -543,10 +543,16 @@ function showUpdateNotification() {
  message.innerHTML = `
    <div class="pwa-update-content">
      <span>Une nouvelle version d'ObsiGate est disponible !</span>
-      <button class="pwa-update-btn" onclick="window.location.reload()">Mettre à jour</button>
-      <button class="pwa-update-dismiss" onclick="this.parentElement.parentElement.remove()">×</button>
+      <button class="pwa-update-btn">Mettre à jour</button>
+      <button class="pwa-update-dismiss">×</button>
    </div>
  `;
+  message.querySelector(".pwa-update-btn").addEventListener("click", function() {
+    window.location.reload();
+  });
+  message.querySelector(".pwa-update-dismiss").addEventListener("click", function() {
+    message.remove();
+  });
  document.body.appendChild(message);

  // Auto-dismiss after 30 seconds
@@ -598,6 +598,17 @@ function applyTheme(themeKey, mode) {
  // Notify other components of theme change
  try {
    root.setAttribute('data-theme', mode);
+    // Syntax highlighting follows the light/dark mode (BUG-078): the theme
+    // *key* is not a mode, so toggling the sheets by key used to disable both
+    // and strip every code block of its colours. Sepia/high-contrast reuse the
+    // light palette.
+    var darkSheet = document.getElementById('hljs-theme-dark');
+    var lightSheet = document.getElementById('hljs-theme-light');
+    if (darkSheet && lightSheet) {
+      var isDark = mode === 'dark';
+      darkSheet.disabled = !isDark;
+      lightSheet.disabled = isDark;
+    }
    document.dispatchEvent(new CustomEvent('themechange', { detail: { theme: themeKey, mode: mode } }));
  } catch(e) {}
 }
@@ -137,14 +137,26 @@ export const RightSidebarManager = {
 // ---------------------------------------------------------------------------
 // Theme
 // ---------------------------------------------------------------------------
+const _THEME_MODES = ["dark", "light", "high-contrast", "sepia"];
+
 export function initTheme() {
-  const saved = localStorage.getItem("obsigate-theme") || "dark";
-  applyTheme(saved);
+  // The theme engine (themes.js) owns the CSS variables and the theme *key*;
+  // the <html data-theme> attribute must carry the *mode* (dark/light/…). Read
+  // the persisted mode here so the first paint and the highlight.js stylesheet
+  // are right before the async theme init runs (BUG-078).
+  let mode = "dark";
+  try { mode = localStorage.getItem("obsigate-theme-mode") || "dark"; } catch (e) { /* ignore */ }
+  applyTheme(mode);
 }

 export function applyTheme(theme) {
-  document.documentElement.setAttribute("data-theme", theme);
-  localStorage.setItem("obsigate-theme", theme);
+  let mode = theme;
+  if (_THEME_MODES.indexOf(mode) === -1) {
+    // Callers may pass a theme key (legacy): fall back to the persisted mode.
+    try { mode = localStorage.getItem("obsigate-theme-mode") || "dark"; } catch (e) { mode = "dark"; }
+  }
+  const isDark = mode === "dark";
+  document.documentElement.setAttribute("data-theme", mode);

  // Update theme button icon and label
  const themeBtn = document.getElementById("theme-toggle");
@@ -152,23 +164,23 @@ export function applyTheme(theme) {
  if (themeBtn && themeLabel) {
    const icon = themeBtn.querySelector("i");
    if (icon) {
-      icon.setAttribute("data-lucide", theme === "dark" ? "moon" : "sun");
+      icon.setAttribute("data-lucide", isDark ? "moon" : "sun");
    }
-    themeLabel.textContent = theme === "dark" ? t('theme.dark') : t('theme.light');
+    themeLabel.textContent = isDark ? t('theme.dark') : t('theme.light');
    safeCreateIcons();
  }

-  // Swap highlight.js theme
+  // Swap highlight.js theme — keyed on the mode, not the theme key (BUG-078).
  const darkSheet = document.getElementById("hljs-theme-dark");
  const lightSheet = document.getElementById("hljs-theme-light");
  if (darkSheet && lightSheet) {
-    darkSheet.disabled = theme !== "dark";
-    lightSheet.disabled = theme !== "light";
+    darkSheet.disabled = !isDark;
+    lightSheet.disabled = isDark;
  }

  // Update Mermaid theme for newly rendered diagrams
  import('./mermaid-viewer.js').then(function(m) {
-    m.updateMermaidTheme(theme === 'dark');
+    m.updateMermaidTheme(isDark);
  }).catch(function() {});
 }

@@ -995,6 +995,284 @@ export function renderVideoViewer(area, data) {
 }


+// ── Excel .xlsx — sheet tabs + editable cells ─────────────────────────────
+// Cells are contenteditable; edits are collected per sheet and sent to
+// PUT /api/file/{vault}/xlsx/save. Formula cells show their text and are
+// never recalculated here.
+//
+// #153 A1/A4 — the read response carries `xlsx_lossy_features` (parts openpyxl
+// drops on save): a banner lists them and the first save asks for an explicit
+// confirmation before retrying with `force: true`. A value starting with "=" or
+// "@" is stored as text unless the user turns the formula toggle on, so a typed
+// `=cmd|…` cannot execute when the file is later opened in Excel.
+// #153 A8 — a sheet bigger than the render caps used to be silently cut: the
+// user saw a short table and no way to tell the rest of the workbook still
+// existed. The backend now reports the real dimensions of every sheet, so the
+// note states exactly what is hidden (and that those cells are not editable
+// here — the workbook itself is untouched). A payload without those fields
+// (older cache) simply shows no note.
+function truncationNote(sheet) {
+  // The cap is the real "shown" figure, not `rows`/`cols`: those are post-trim
+  // (a sparse sheet renders 1x1) while the note must say how far the view
+  // reaches.
+  const cap = { rows: Number(sheet.max_rows) || 0, cols: Number(sheet.max_cols) || 0 };
+  const reasons = [];
+  if (Number(sheet.total_rows) > cap.rows) {
+    reasons.push(t("xlsx.truncated_rows", { shown: cap.rows, total: Number(sheet.total_rows) }));
+  }
+  if (Number(sheet.total_cols) > cap.cols) {
+    reasons.push(t("xlsx.truncated_cols", { shown: cap.cols, total: Number(sheet.total_cols) }));
+  }
+  if (!reasons.length) return "";
+  return `<div class="xlsx-truncated" role="note">
+      <i data-lucide="scissors" class="xlsx-truncated-icon"></i>
+      <div class="xlsx-warning-body">
+        <strong>${escapeHtml(t("xlsx.truncated_title"))}</strong>
+        <span>${escapeHtml(reasons.join(" "))}</span>
+        <span class="xlsx-warning-hint">${escapeHtml(t("xlsx.truncated_hint"))}</span>
+      </div>
+    </div>`;
+}
+
+export function renderXlsxViewer(area, data) {
+  const sheets = data.xlsx_sheets || [];
+  const lossy = data.xlsx_lossy_features || [];
+  // Session-scoped state: once the lossy write is confirmed, the rest of the
+  // session saves without asking again (never persisted — a confirmation is
+  // per workbook, not a global preference).
+  let lossyConfirmed = false;
+  let allowFormula = false;
+  const tabs = sheets.length > 1
+    ? `<div class="xlsx-tabs">${sheets.map((s, i) =>
+        `<button class="xlsx-tab${i === 0 ? " active" : ""}" data-sheet="${i}">${escapeHtml(s.name)}</button>`
+      ).join("")}</div>`
+    : "";
+  const panels = sheets.map((s, i) =>
+    `<div class="xlsx-panel" data-sheet="${i}"${i === 0 ? "" : ' style="display:none"'}>${truncationNote(s)}${s.html}</div>`
+  ).join("");
+  const lossWarning = lossy.length
+    ? `<div class="xlsx-warning" role="note">
+         <i data-lucide="alert-triangle" class="xlsx-warning-icon"></i>
+         <div class="xlsx-warning-body">
+           <strong>${escapeHtml(t("xlsx.lossy_title"))}</strong>
+           <span class="xlsx-warning-list">${lossy
+             .map((f) => `<span class="xlsx-warning-tag">${escapeHtml(t("xlsx.feature_" + f))}</span>`)
+             .join("")}</span>
+           <span class="xlsx-warning-hint">${escapeHtml(t("xlsx.lossy_hint"))}</span>
+         </div>
+       </div>`
+    : "";
+
+  area.innerHTML = `
+    <div class="xlsx-viewer">
+      <div class="xlsx-toolbar">
+        ${tabs}
+        <span class="xlsx-toolbar-actions">
+          <button class="btn-action xlsx-formula-toggle" id="xlsx-formula-btn" type="button"
+                  aria-pressed="false" title="${escapeHtml(t("xlsx.formula_toggle_title"))}">f(x)</button>
+          <button class="btn-action" id="xlsx-save-btn" disabled>${t("common.save")}</button>
+          <button class="btn-action" id="xlsx-download-btn">
+            <i data-lucide="download" style="width:14px;height:14px"></i> ${t("viewer.download")}
+          </button>
+        </span>
+      </div>
+      ${lossWarning}
+      <div class="xlsx-panels">${panels}</div>
+    </div>`;
+
+  const saveBtn = area.querySelector("#xlsx-save-btn");
+  const panelEls = [...area.querySelectorAll(".xlsx-panel")];
+  const dirtyCount = () => area.querySelectorAll("td.xlsx-dirty").length;
+  const refreshSaveState = () => { saveBtn.disabled = dirtyCount() === 0; };
+
+  // #153 A9bis — the first render stops at MAX_ROWS/MAX_COLS; the tail is
+  // fetched window by window from GET …/xlsx/sheet when the user reaches the
+  // end of a truncated sheet (scroll sentinel) or clicks « Charger la suite ».
+  // Appended rows reuse the exact same edit pipeline as the initial render.
+  const wireLazyRows = (panel) => {
+    const meta = sheets[Number(panel.dataset.sheet)] || {};
+    if (!meta.truncated) return;
+    const wrapper = panel.querySelector(".csv-table-wrapper");
+    const table = panel.querySelector(".xlsx-table tbody");
+    if (!wrapper || !table) return;
+
+    let offset = Number(meta.rows) || 0;
+    const total = Number(meta.total_rows) || 0;
+    let loading = false;
+    let done = offset >= total;
+
+    const foot = document.createElement("div");
+    foot.className = "xlsx-load-more";
+    const refreshFoot = () => {
+      foot.textContent = done
+        ? ""
+        : `${t("xlsx.truncated_rows", { shown: offset, total })} · ${t("xlsx.load_more")}`;
+      foot.classList.toggle("done", done);
+    };
+    refreshFoot();
+    foot.addEventListener("click", () => { if (!done) loadMore(); });
+    wrapper.insertAdjacentElement("afterend", foot);
+
+    const appendWindow = (win) => {
+      const doc = new DOMParser().parseFromString(`<table>${win.html}</table>`, "text/html");
+      // No `tbody` selector: the fragment embeds its own wrapper div, so the
+      // parse yields bare `<tr>` inside `<table>` (the rows we want) — while
+      // any `tbody` in the fragment belongs to the *embedded* wrapper table.
+      doc.querySelectorAll("tr").forEach((tr) => table.appendChild(tr));
+      offset = win.offset + win.rows;
+      done = !win.has_more;
+      refreshFoot();
+      // The new rows must behave like the initial ones: contenteditable,
+      // dirty tracking, cached-value tooltip, icons in the fresh footnote.
+      panel.querySelectorAll("tbody tr:not([data-wired]) td").forEach(setupCell);
+      panel.querySelectorAll("tbody tr").forEach((tr) => tr.setAttribute("data-wired", "1"));
+      const cachedEls = panel.querySelectorAll(".xlsx-cached[data-cached-value]");
+      cachedEls.forEach((el) => { if (!el.title) el.title = t("xlsx.cached_value_title"); });
+      safeCreateIcons();
+    };
+
+    const loadMore = async () => {
+      if (loading || done) return;
+      loading = true;
+      const label = foot.textContent;
+      foot.textContent = t("xlsx.loading_more");
+      try {
+        const win = await api(
+          `/api/file/${encodeURIComponent(data.vault)}/xlsx/sheet?path=${encodeURIComponent(data.path)}` +
+          `&sheet=${encodeURIComponent(sheets[Number(panel.dataset.sheet)].name)}` +
+          `&offset=${offset}&limit=500`,
+        );
+        appendWindow(win);
+      } catch (err) {
+        foot.textContent = label; // restore: a failed fetch must not eat the button
+        showToast(`${t("xlsx.load_error")}: ${err.message || err}`, "error");
+      } finally {
+        loading = false;
+      }
+    };
+
+    // Scroll sentinel: near the bottom of the wrapper, pull the next window.
+    wrapper.addEventListener("scroll", () => {
+      if (loading || done) return;
+      if (wrapper.scrollTop + wrapper.clientHeight >= wrapper.scrollHeight - 120) loadMore();
+    });
+  };
+
+  const setupCell = (td) => {
+    td.contentEditable = "true";
+    td.spellcheck = false;
+    td.dataset.orig = td.textContent;
+    td.addEventListener("input", () => {
+      td.classList.add("xlsx-dirty");
+      refreshSaveState();
+    });
+    td.addEventListener("keydown", (e) => {
+      if (e.key === "Enter") { e.preventDefault(); td.blur(); }
+      if (e.key === "Escape") {
+        td.textContent = td.dataset.orig;
+        td.classList.remove("xlsx-dirty");
+        refreshSaveState();
+      }
+    });
+    td.addEventListener("paste", (e) => {
+      e.preventDefault();
+      const text = (e.clipboardData || window.clipboardData).getData("text").replace(/\r?\n/g, " ");
+      document.execCommand("insertText", false, text);
+    });
+  };
+
+  // #153 A12 — the backend marks the last value Excel computed; the wording is
+  // translated here so the tooltip follows the UI language.
+  area.querySelectorAll(".xlsx-cached[data-cached-value]").forEach((el) => {
+    el.title = t("xlsx.cached_value_title");
+  });
+
+  // Editable cells: Enter blurs, Escape reverts, paste stays single-line.
+  area.querySelectorAll(".xlsx-table td").forEach(setupCell);
+  panelEls.forEach(wireLazyRows);
+
+  area.querySelectorAll(".xlsx-tab").forEach((tab) => {
+    tab.addEventListener("click", () => {
+      const idx = tab.dataset.sheet;
+      area.querySelectorAll(".xlsx-tab").forEach((x) => x.classList.toggle("active", x === tab));
+      panelEls.forEach((p) => { p.style.display = p.dataset.sheet === idx ? "" : "none"; });
+    });
+  });
+
+  // Formula toggle (#153 A4) — opt-in for this viewing session only.
+  const formulaBtn = area.querySelector("#xlsx-formula-btn");
+  formulaBtn.addEventListener("click", () => {
+    allowFormula = !allowFormula;
+    formulaBtn.setAttribute("aria-pressed", String(allowFormula));
+    formulaBtn.classList.toggle("active", allowFormula);
+  });
+
+  const putSheet = (job, force) => api(
+    `/api/file/${encodeURIComponent(data.vault)}/xlsx/save?path=${encodeURIComponent(data.path)}`,
+    {
+      method: "PUT",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify({ ...job, allow_formula: allowFormula, force }),
+    },
+  );
+
+  saveBtn.addEventListener("click", async () => {
+    // One PUT per sheet (dirty cells can span tabs before a save).
+    const jobs = panelEls
+      .map((panel) => {
+        const cells = {};
+        panel.querySelectorAll("td.xlsx-dirty").forEach((td) => { cells[td.dataset.cell] = td.textContent; });
+        return { sheet: sheets[Number(panel.dataset.sheet)].name, cells };
+      })
+      .filter((job) => Object.keys(job.cells).length);
+    if (!jobs.length) return;
+    saveBtn.disabled = true;
+    try {
+      for (const job of jobs) {
+        // 409 xlsx_lossy_content → confirm once, then retry with force: true.
+        // (Also covers a workbook that became lossy while it was open.)
+        let force = lossyConfirmed;
+        for (;;) {
+          try {
+            await putSheet(job, force);
+            break;
+          } catch (err) {
+            if (err && err.code === "xlsx_lossy_content" && !lossyConfirmed) {
+              const features = (err.details && err.details.features) || lossy;
+              const labels = features.map((f) => t("xlsx.feature_" + f)).join(", ");
+              if (!confirm(t("xlsx.lossy_confirm", { features: labels }))) throw err;
+              lossyConfirmed = true;
+              force = true;
+              continue;
+            }
+            throw err;
+          }
+        }
+      }
+      area.querySelectorAll("td.xlsx-dirty").forEach((td) => {
+        td.classList.remove("xlsx-dirty");
+        td.dataset.orig = td.textContent;
+      });
+      refreshSaveState();
+      showToast(t("editor.saved"), "success");
+    } catch (err) {
+      refreshSaveState();
+      // A refused confirmation is a decision, not a failure: neutral toast.
+      if (err && err.code === "xlsx_lossy_content") {
+        showToast(t("xlsx.lossy_cancelled"), "info");
+      } else {
+        showToast(`${t("editor.save_error")}: ${err.message || err}`, "error");
+      }
+    }
+  });
+
+  area.querySelector("#xlsx-download-btn").addEventListener("click", () => {
+    window.open(`/api/file/${encodeURIComponent(data.vault)}/download?path=${encodeURIComponent(data.path)}`, "_blank");
+  });
+
+  safeCreateIcons();
+}
+
 export function renderFile(data) {
  // #93 — An inline edition session (#editor-container mounted in the content
  // area) is destroyed by this very re-render: release it first so the editor
@@ -1036,10 +1314,10 @@ export function renderFile(data) {
      <div class="pdf-viewer-container">
        <div class="pdf-toolbar">
          <span class="pdf-info">PDF — ${pages} pages</span>
-          <button class="btn-action" onclick="window.open('${pdfUrl}', '_blank')">
+          <button class="btn-action" data-pdf-url="${pdfUrl}">
            <i data-lucide="external-link" style="width:14px;height:14px"></i> Plein écran
          </button>
-          <button class="btn-action" onclick="window.open('/api/file/${encodeURIComponent(data.vault)}/download?path=${encodeURIComponent(data.path)}', '_blank')">
+          <button class="btn-action" data-download-url="/api/file/${encodeURIComponent(data.vault)}/download?path=${encodeURIComponent(data.path)}">
            <i data-lucide="download" style="width:14px;height:14px"></i> Télécharger
          </button>
        </div>
@@ -1048,6 +1326,11 @@ export function renderFile(data) {
          <iframe src="${pdfUrl}" data-pdf-url="${pdfUrl}" class="pdf-iframe" title="${escapeHtml(data.title)}"></iframe>
        </div>
      </div>`;
+    area.querySelectorAll('.pdf-toolbar .btn-action').forEach((btn) => {
+      btn.addEventListener('click', () => {
+        window.open(btn.dataset.pdfUrl || btn.dataset.downloadUrl, '_blank');
+      });
+    });
    area.querySelectorAll('.pdf-toc a[data-page]').forEach((link) => {
      link.addEventListener('click', (e) => {
        e.preventDefault();
@@ -1058,6 +1341,12 @@ export function renderFile(data) {
    return;
  }

+  // Handle Excel .xlsx — editable table view (display / edit / download)
+  if (data.is_xlsx) {
+    renderXlsxViewer(area, data);
+    return;
+  }
+
  // Handle Excalidraw — render in iframe editor
  if (data.is_excalidraw) {
    renderExcalidraw(area, data, data.vault, data.path);
@@ -1418,7 +1707,12 @@ export function renderFile(data) {
  // Assemble
  area.innerHTML = "";
  area.appendChild(breadcrumb);
-  area.appendChild(el("div", { class: "file-header" }, [el("div", { class: "file-title" }, [document.createTextNode(data.title)]), tagsDiv, el("div", { class: "file-actions" }, fileActions)]));
+  area.appendChild(el("div", { class: "file-header" }, [el("div", { class: "file-title" }, [document.createTextNode(data.title)]), tagsDiv]));
+  // #115 — the action bar is a direct child of the scroll container (wrapped in
+  // `.file-toolbar`) so it can stay pinned while long documents scroll; nested
+  // inside `.file-header` it would stop sticking as soon as the header left the
+  // viewport.
+  area.appendChild(el("div", { class: "file-toolbar" }, [el("div", { class: "file-actions" }, fileActions)]));
  if (fmSection) area.appendChild(fmSection);
  area.appendChild(mdDiv);
  area.appendChild(rawDiv);
@@ -80,12 +80,16 @@
  "admin.users_title": "User management",
  "ai.action_applied": "Applied ✓",
  "ai.action_apply": "Apply",
+  "ai.action_apply_all": "Approve all ({count})",
  "ai.action_applying": "Applying…",
  "ai.action_create_dir": "Create folder {path}",
  "ai.action_create_file": "Create file {path}",
  "ai.action_created_dir": "Folder created: {path}",
  "ai.action_created_file": "File created: {path}",
  "ai.action_failed": "Action failed: {error}",
+  "ai.confirm_actions": "{count} actions to approve",
+  "ai.stop": "Stop the assistant",
+  "ai.stopped": "Run stopped.",
  "ai.agent_mode_off": "Agent mode off (read/search + actions)",
  "ai.agent_mode_on": "Agent mode on (read/search tools + actions)",
  "ai.casual": "Casual tone",
@@ -382,6 +386,7 @@
  "config.key_deleted": "Key deleted:",
  "config.avatar_label": "Profile picture",
  "config.avatar_hint": "PNG, JPG or WEBP — square image cropped and resized to 256 px. Shown in the sidebar.",
+  "config.avatar_presets_label": "Or pick a preset avatar",
  "config.avatar_choose": "Choose an image",
  "config.avatar_remove": "Remove picture",
  "config.avatar_updated": "Profile picture updated",
@@ -578,6 +583,7 @@
  "config.test": "Test",
  "config.timeout_label": "Search timeout (ms)",
  "config.title": "Settings",
+  "config.toc_close": "Close contents",
  "config.toc_toggle": "Show contents",
  "config.title_boost": "Title boost",
  "config.title_boost_hint": "Relevance multiplier for title matches",
@@ -1351,6 +1357,7 @@
  "help.assistant_links": "Cited files and paths are links: a bare filename copies the name to the clipboard, a folder is revealed in the tree, and a file path opens it in the viewer.",
  "help.assistant_sessions": "The header history icon lists past sessions (reopen or delete); “+” starts a new conversation.",
  "help.assistant_agent": "The \"agent mode\" button enables tools (read, list, search); modifying actions require confirmation with a change preview.",
+  "help.assistant_agent_run": "Actions appear in the thread: the assistant groups modifications into a single approval (\"Approve all\") and refreshes the file tree and the open document as soon as they are applied. The send button becomes \"Stop\" to interrupt the run at any time.",
  "help.assistant_resize": "The left edge of the panel is resizable; the width is remembered.",
  "help.assistant_at": "Type @ to attach a file or directory to the context, or to attach an image from a directory.",
  "help.assistant_slash": "Type / to run a skill (research, summary, correction, plan…) or an admin command (/help, /providers, /model, /keys).",
@@ -1631,25 +1638,21 @@
  "search.whole_word": "Whole word",
  "settings.about": "📦 About",
  "settings.ai": "🤖 AI",
-  "settings.backend": "⚙️ Backend",
-  "settings.backend_hint": "These settings are saved on the server. Some require a restart or reindexing.",
  "settings.backend_section": "Backend Settings",
  "settings.client_label": "These settings apply immediately on the client side.",
  "settings.diagnostics": "🩺 Diagnostics",
+  "settings.explorer": "Files",
  "settings.history_count_desc": "Between 5 and 100 files (saved on server)",
  "settings.history_count_label": "Files in history",
  "settings.history_section": "Recent History",
  "settings.no_restart_badge": "No restart needed",
  "settings.profile": "👤 Profile",
  "settings.reindex": "Force reindex",
-  "settings.restart_badge": "Restart required",
-  "settings.save": "Save",
  "settings.search": "Search",
-  "settings.tabs": "Tabs",
  "settings.search_placeholder": "Search...",
  "settings.sync": "🔄 Sync",
+  "settings.tabs": "Tabs",
  "settings.themes": "🎨 Themes",
-  "settings.plugins": "🧩 Plugins",
  "share.copied": "Link copied!",
  "share.copy_link": "Copy link",
  "share.create": "Create share link",
@@ -1820,6 +1823,27 @@
  "viewer.copy": "Copy",
  "viewer.copy_error": "Copy error",
  "viewer.download": "Download",
+  "xlsx.lossy_title": "Simplified save",
+  "xlsx.lossy_hint": "ObsiGate cannot preserve these elements: saving will ask for your confirmation.",
+  "xlsx.lossy_confirm": "Save anyway? The following will be lost: {features}",
+  "xlsx.lossy_cancelled": "Save cancelled",
+  "xlsx.formula_toggle_title": "Treat “=” and “@” as formulas (off by default)",
+  "xlsx.cached_value_title": "Last value calculated by Excel",
+  "xlsx.truncated_title": "Truncated sheet",
+  "xlsx.truncated_rows": "{shown} of {total} rows displayed.",
+  "xlsx.truncated_cols": "{shown} of {total} columns displayed.",
+  "xlsx.truncated_hint": "Cells outside the displayed area cannot be edited here; the workbook is unchanged.",
+  "xlsx.load_more": "Load more",
+  "xlsx.loading_more": "Loading…",
+  "xlsx.load_error": "Could not load the remaining rows",
+  "xlsx.feature_cached_values": "cached values",
+  "xlsx.feature_slicers": "slicers and timelines",
+  "xlsx.feature_form_controls": "form controls",
+  "xlsx.feature_connections": "connections and queries",
+  "xlsx.feature_custom_xml": "custom XML",
+  "xlsx.feature_signature": "digital signature",
+  "xlsx.feature_rich_comments": "rich comments",
+  "xlsx.feature_macros": "macros",
  "viewer.download_md": "Download as .md",
  "viewer.download_file": "Download file",
  "viewer.pretty": "Pretty",
@@ -1984,6 +2008,10 @@
  "ai.step.git_issues": "Searched issues: {value}",
  "ai.step.git_file": "Read a repo file: {value}",
  "ai.step.xlsx_create": "Spreadsheet proposed: {value}",
+  "ai.step.xlsx_sheets": "Workbook sheets listed: {value}",
+  "ai.step.xlsx_read": "Workbook read: {value}",
+  "ai.step.xlsx_update": "Cells edited: {value}",
+  "ai.step.xlsx_append": "Rows appended: {value}",
  "ai.step.docx_create": "Word document proposed: {value}",
  "ai.step.csv_create": "CSV file proposed: {value}",
  "ai.step.pdf_create": "PDF document proposed: {value}",
@@ -80,12 +80,16 @@
  "admin.users_title": "Gestion utilisateurs",
  "ai.action_applied": "Appliqué ✓",
  "ai.action_apply": "Appliquer",
+  "ai.action_apply_all": "Tout approuver ({count})",
  "ai.action_applying": "Application…",
  "ai.action_create_dir": "Créer le dossier {path}",
  "ai.action_create_file": "Créer le fichier {path}",
  "ai.action_created_dir": "Dossier créé : {path}",
  "ai.action_created_file": "Fichier créé : {path}",
  "ai.action_failed": "Échec de l'action : {error}",
+  "ai.confirm_actions": "{count} actions à approuver",
+  "ai.stop": "Arrêter l'assistant",
+  "ai.stopped": "Exécution arrêtée.",
  "ai.agent_mode_off": "Mode agent désactivé (lecture/recherche + actions)",
  "ai.agent_mode_on": "Mode agent activé (outils de lecture/recherche + actions)",
  "ai.casual": "Ton décontracté",
@@ -382,6 +386,7 @@
  "config.key_deleted": "Clé supprimée :",
  "config.avatar_label": "Photo de profil",
  "config.avatar_hint": "PNG, JPG ou WEBP — image carrée recadrée et réduite à 256 px. Apparaît dans la barre latérale.",
+  "config.avatar_presets_label": "Ou choisissez un avatar prédéfini",
  "config.avatar_choose": "Choisir une image",
  "config.avatar_remove": "Supprimer la photo",
  "config.avatar_updated": "Photo de profil mise à jour",
@@ -578,6 +583,7 @@
  "config.test": "Tester",
  "config.timeout_label": "Timeout recherche (ms)",
  "config.title": "Configuration",
+  "config.toc_close": "Fermer le sommaire",
  "config.toc_toggle": "Afficher le sommaire",
  "config.title_boost": "Boost titre",
  "config.title_boost_hint": "Multiplicateur de pertinence pour les correspondances dans le titre",
@@ -1351,6 +1357,7 @@
  "help.assistant_links": "Les fichiers et chemins cités sont des liens : un simple nom de fichier copie le nom dans le presse-papiers, un dossier est révélé dans l'arborescence, et un chemin de fichier l'ouvre dans le viewer.",
  "help.assistant_sessions": "L'icône historique de l'en-tête liste les sessions passées (recharger ou supprimer) ; « + » démarre une nouvelle conversation.",
  "help.assistant_agent": "Le bouton « mode agent » active les outils (lire, lister, chercher) ; les actions de modification demandent une confirmation avec aperçu des changements.",
+  "help.assistant_agent_run": "Les actions s'affichent dans le fil : l'assistant regroupe les modifications en une seule approbation (« Tout approuver ») et met à jour l'arborescence et le document ouvert dès qu'elles sont appliquées. Le bouton d'envoi devient « Stop » pour interrompre l'exécution à tout moment.",
  "help.assistant_resize": "Le bord gauche du panneau est redimensionnable ; la largeur est mémorisée.",
  "help.assistant_at": "Tapez @ pour joindre un fichier ou un répertoire au contexte, ou pour attacher une image d'un répertoire.",
  "help.assistant_slash": "Tapez / pour lancer un skill (recherche, résumé, correction, plan…) ou une commande admin (/help, /providers, /model, /keys).",
@@ -1631,25 +1638,21 @@
  "search.whole_word": "Mot entier",
  "settings.about": "📦 À propos",
  "settings.ai": "🤖 IA",
-  "settings.backend": "⚙️ Backend",
-  "settings.backend_hint": "Ces paramètres sont sauvegardés sur le serveur. Certains nécessitent un redémarrage ou une réindexation.",
  "settings.backend_section": "Paramètres backend",
  "settings.client_label": "Ces paramètres s'appliquent immédiatement côté client.",
  "settings.diagnostics": "🩺 Diagnostics",
+  "settings.explorer": "Explorateur",
  "settings.history_count_desc": "Entre 5 et 100 fichiers (sauvegarde sur le serveur)",
  "settings.history_count_label": "Nombre de fichiers dans l'historique",
  "settings.history_section": "Historique récent",
  "settings.no_restart_badge": "Redémarrage non requis",
  "settings.profile": "👤 Profil",
  "settings.reindex": "Forcer réindexation",
-  "settings.restart_badge": "Redémarrage requis",
-  "settings.save": "Sauvegarder",
  "settings.search": "Recherche",
-  "settings.tabs": "Tabs",
  "settings.search_placeholder": "Rechercher...",
  "settings.sync": "🔄 Synchronisation",
+  "settings.tabs": "Onglets",
  "settings.themes": "🎨 Thèmes",
-  "settings.plugins": "🧩 Plugins",
  "share.copied": "Lien copié !",
  "share.copy_link": "Copier le lien",
  "share.create": "Créer un lien de partage",
@@ -1820,6 +1823,27 @@
  "viewer.copy": "Copier",
  "viewer.copy_error": "Erreur lors de la copie",
  "viewer.download": "Télécharger",
+  "xlsx.lossy_title": "Enregistrement simplifié",
+  "xlsx.lossy_hint": "Ces éléments ne peuvent pas être conservés par ObsiGate : une sauvegarde vous demandera confirmation.",
+  "xlsx.lossy_confirm": "Enregistrer quand même ? Les éléments suivants seront perdus : {features}",
+  "xlsx.lossy_cancelled": "Sauvegarde annulée",
+  "xlsx.formula_toggle_title": "Interpréter « = » et « @ » comme des formules (désactivé par défaut)",
+  "xlsx.cached_value_title": "Dernière valeur calculée par Excel",
+  "xlsx.truncated_title": "Feuille tronquée",
+  "xlsx.truncated_rows": "{shown} lignes affichées sur {total}.",
+  "xlsx.truncated_cols": "{shown} colonnes affichées sur {total}.",
+  "xlsx.truncated_hint": "Les cellules hors de l'affichage ne sont pas éditables ici ; le classeur n'est pas modifié.",
+  "xlsx.load_more": "Charger la suite",
+  "xlsx.loading_more": "Chargement…",
+  "xlsx.load_error": "Chargement de la suite impossible",
+  "xlsx.feature_cached_values": "valeurs calculées",
+  "xlsx.feature_slicers": "segments et chronologies",
+  "xlsx.feature_form_controls": "contrôles de formulaire",
+  "xlsx.feature_connections": "connexions et requêtes",
+  "xlsx.feature_custom_xml": "XML personnalisé",
+  "xlsx.feature_signature": "signature numérique",
+  "xlsx.feature_rich_comments": "commentaires enrichis",
+  "xlsx.feature_macros": "macros",
  "viewer.download_md": "Télécharger en .md",
  "viewer.download_file": "Télécharger le fichier",
  "viewer.pretty": "Pretty",
@@ -1984,6 +2008,10 @@
  "ai.step.git_issues": "Issues recherchées : {value}",
  "ai.step.git_file": "Fichier de dépôt lu : {value}",
  "ai.step.xlsx_create": "Tableur proposé : {value}",
+  "ai.step.xlsx_sheets": "Feuilles du classeur listées : {value}",
+  "ai.step.xlsx_read": "Classeur lu : {value}",
+  "ai.step.xlsx_update": "Cellules modifiées : {value}",
+  "ai.step.xlsx_append": "Lignes ajoutées : {value}",
  "ai.step.docx_create": "Document Word proposé : {value}",
  "ai.step.csv_create": "Fichier CSV proposé : {value}",
  "ai.step.pdf_create": "Document PDF proposé : {value}",
@@ -38,7 +38,7 @@
      }
    })();
  </script>
-  <link rel="stylesheet" href="/static/style.css?v=2">
+  <link rel="stylesheet" href="/static/style.css?v=3">
  <script src="https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.min.js"></script>
  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/github-dark.min.css" id="hljs-theme-dark">
  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/github.min.css" id="hljs-theme-light" disabled>
@@ -871,8 +871,13 @@ const RightSidebarManager = {
            group.forEach(function (b) { actionsDiv.appendChild(b); });
          });

-        header.appendChild(actionsDiv);
        area.appendChild(header);
+        // #115 — pinned action bar (direct child of the scroll container) so it
+        // stays visible while long documents scroll, as in the main viewer.
+        const toolbar = document.createElement("div");
+        toolbar.className = "file-toolbar";
+        toolbar.appendChild(actionsDiv);
+        area.appendChild(toolbar);

        // Frontmatter — Accent Card
        if (data.frontmatter && Object.keys(data.frontmatter).length > 0) {
@@ -1767,7 +1767,22 @@ select {

 /* File header */
 .file-header {
-  margin-bottom: 20px;
+  margin-bottom: 8px;
+}
+/* #115 — sticky document action bar: stays visible while long content scrolls.
+   A direct child of `.content-area` (the scroll container), opaque so the text
+   scrolling underneath never bleeds through. */
+.file-toolbar {
+  position: sticky;
+  top: 0;
+  z-index: 30;
+  margin: 0 0 16px;
+  padding: 8px 0;
+  background: var(--bg-primary);
+  border-bottom: 1px solid var(--border);
+}
+.file-toolbar .file-actions {
+  margin-top: 0;
 }
 .file-title {
  font-family: "JetBrains Mono", monospace;
@@ -2670,6 +2685,32 @@ select {
 .profile-avatar-remove-btn { background: var(--danger-bg) !important; color: var(--danger) !important; border-color: var(--danger) !important; }
 .profile-avatar-error { font-size: 0.7rem; color: var(--danger); margin-top: 8px; }
 .profile-avatar-error.hidden { display: none; }
+/* #117 — preset avatar gallery */
+.profile-avatar-presets {
+  display: grid;
+  grid-template-columns: repeat(6, minmax(0, 1fr));
+  gap: 8px;
+  margin-top: 10px;
+  max-width: 420px;
+}
+.profile-avatar-preset {
+  padding: 0;
+  margin: 0;
+  border: 2px solid var(--border);
+  border-radius: 50%;
+  overflow: hidden;
+  cursor: pointer;
+  background: var(--bg-secondary);
+  aspect-ratio: 1;
+  transition: border-color 0.15s ease, box-shadow 0.15s ease, transform 0.1s ease;
+}
+.profile-avatar-preset img { width: 100%; height: 100%; object-fit: cover; display: block; }
+.profile-avatar-preset:hover { border-color: var(--accent); transform: translateY(-1px); }
+.profile-avatar-preset.active {
+  border-color: var(--accent);
+  box-shadow: 0 0 0 3px color-mix(in srgb, var(--accent) 22%, transparent);
+}
+.profile-avatar-preset:disabled { opacity: 0.6; cursor: progress; }
 .config-save-btn {
  padding: 6px 18px; font-size: 0.75rem; font-weight: 600;
  background: var(--accent); color: #fff;
@@ -3164,6 +3205,23 @@ select {
  margin-right: 8px;
 }
 .help-hamburger:hover { color: var(--accent); border-color: var(--accent); }
+/* #114: close button for the mobile TOC drawer (hidden on desktop where the
+   nav is always visible). Scoped to #config-modal so the help modal — which
+   has no such button — is unaffected. */
+.help-toc-close {
+  display: none;
+  width: 44px;
+  height: 44px;
+  padding: 0;
+  border: 1px solid var(--border);
+  border-radius: 6px;
+  background: transparent;
+  color: var(--text-secondary);
+  cursor: pointer;
+  align-items: center;
+  justify-content: center;
+  flex-shrink: 0;
+}
 .help-nav {
  width: 260px;
  min-width: 260px;
@@ -4777,22 +4835,209 @@ body.resizing-v {
  }
 }

-/* BUG-071: Configurations modal — mobile usability (viewport ≤ 768px).
-   #config-nav shares the .help-nav rule that hides it, but the config modal
-   had no toggle (unlike the help modal): the header hamburger
-   (#config-hamburger, same .help-hamburger treatment) reveals it as a
-   collapsible top block. Two-column grids and fixed-width add-rows are
-   stacked/wrapped so nothing overflows a 360px viewport. */
+/* --- #114: Configurations modal — mobile UX refonte (viewport ≤ 768px).
+   Builds on BUG-071 (which stacked grids/wrapped rows): the TOC becomes a
+   left slide-over drawer with a backdrop, the modal goes full-screen with
+   100dvh, every interactive control gets a ≥44px touch target, inputs are
+   ≥16px to stop iOS auto-zoom, and the backend Save row sticks to the
+   bottom of its scroll container. The drawer is opened by JS toggling
+   .config-toc-open on #config-modal (backdrop + nav visibility) plus an
+   inline display on #config-nav (kept for the display-reset contract). */
@media (max-width: 768px) {
-  /* TOC as a collapsible top block (JS toggles inline display flex/none,
-     which wins over the hiding rule); the list scrolls within a capped nav. */
-  #config-modal #config-nav {
+  /* Full-screen modal with dynamic viewport height (mobile browser chrome). */
+  #config-modal {
+    padding: 0;
+    align-items: stretch;
+  }
+  #config-modal .editor-container {
+    max-width: 100%;
+    width: 100%;
+    height: 100vh;
+    height: 100dvh;
+    max-height: 100vh;
+    max-height: 100dvh;
+    border-radius: 0;
+    border: none;
+    overflow: hidden;
+  }
+  /* Hamburger: 44px touch target (shared .help-hamburger rule in the help
+     modal media query also sets 44px; this scopes it to the config modal
+     in case rule order changes). */
+  #config-modal .help-hamburger {
+    display: inline-flex;
+    width: 44px;
+    height: 44px;
+    min-width: 44px;
+  }
+  /* Close button in the TOC drawer header. */
+  #config-modal .help-toc-close {
+    display: inline-flex;
+    align-items: center;
+    justify-content: center;
+  }
+  #config-modal .help-nav-header {
+    padding: 12px 12px 8px;
+    gap: 8px;
+  }
+  /* Close (X) in the modal header. */
+  #config-modal .editor-btn {
+    width: 44px;
+    height: 44px;
+    min-width: 44px;
+  }
+  /* Anti-zoom: ≥16px font on text fields inside the config modal. Never
+     target input[type=text] globally — it would clobber .mfa-code-input. */
+  #config-modal .config-input,
+  #config-modal .config-select,
+  #config-modal .help-nav-search,
+  #config-modal .profile-field .config-input,
+  #config-modal .profile-field .config-select {
+    min-height: 44px;
+    font-size: 16px;
+  }
+  #config-modal .config-input--num {
+    width: 100%;
+    text-align: left;
+  }
+  #config-modal .help-nav-search-wrap {
+    padding: 0 12px 10px;
+  }
+  #config-modal .help-search-clear {
+    min-width: 44px;
+    min-height: 44px;
+  }
+  #config-modal .help-nav-link {
+    min-height: 44px;
+    display: flex;
+    align-items: center;
+    box-sizing: border-box;
+  }
+  /* Sticky Save row inside #cfg-backend-settings (its scroll container is
+     #config-scroll.help-content, which is the overflow ancestor). */
+  #config-modal .config-actions-row {
+    position: sticky;
+    bottom: 0;
+    z-index: 5;
+    display: flex;
+    flex-direction: column;
+    gap: 8px;
+    margin-top: 16px;
+    margin-bottom: 0;
+    padding: 10px 0 calc(10px + env(safe-area-inset-bottom, 0));
+    background: var(--bg-primary);
+    border-top: 1px solid var(--border);
+  }
+  #config-modal .config-actions-row .config-btn-save,
+  #config-modal .config-actions-row .config-btn-secondary,
+  #config-modal .config-actions-row .config-btn-primary,
+  #config-modal .config-actions-row .config-btn-danger {
+    flex: 1 1 auto;
+    width: 100%;
+    min-height: 44px;
+    box-sizing: border-box;
+  }
+  #config-modal .config-btn-save,
+  #config-modal .config-btn-secondary,
+  #config-modal .config-btn-primary,
+  #config-modal .config-btn-danger,
+  #config-modal .config-btn-add,
+  #config-modal .config-btn-sm,
+  #config-modal .mfa-link-btn,
+  #config-modal .theme-action-btn,
+  #config-modal .profile-avatar-actions .config-btn-secondary,
+  #config-modal .editor-btn {
+    min-height: 44px;
+    box-sizing: border-box;
+  }
+  #config-modal .config-btn-sm,
+  #config-modal .mfa-link-btn,
+  #config-modal .theme-action-btn {
+    min-width: 44px;
+    padding-left: 12px;
+    padding-right: 12px;
+  }
+  /* Profile avatar row: wrap instead of overflowing a 360px viewport. */
+  #config-modal .profile-avatar-row {
+    flex-wrap: wrap;
+    gap: 12px;
+  }
+  /* Preset avatars: 4 per row on narrow screens (44px touch targets). */
+  #config-modal .profile-avatar-presets {
+    grid-template-columns: repeat(4, minmax(0, 1fr));
+    max-width: 100%;
+  }
+  #config-modal .profile-form {
+    max-width: 100%;
+  }
+  /* MFA: one-column recovery list, wrapping action rows, full-width code
+     input with a readable letter-spacing (12px overflows 360px). */
+  #config-modal .mfa-recovery-list {
+    grid-template-columns: 1fr;
+    gap: 6px;
+  }
+  #config-modal .mfa-verify-section,
+  #config-modal .mfa-recovery-actions,
+  #config-modal .mfa-disable-actions {
+    flex-wrap: wrap;
+  }
+  #config-modal .mfa-verify-section .config-input,
+  #config-modal .mfa-verify-section .config-btn-primary {
+    flex: 1 1 100%;
    width: 100%;
    min-width: 0;
-    max-width: 100%;
-    border-right: none;
-    border-bottom: 1px solid var(--border);
-    max-height: 46vh;
+    box-sizing: border-box;
+  }
+  #config-modal .mfa-recovery-actions .config-btn-secondary,
+  #config-modal .mfa-disable-actions .config-btn-danger {
+    flex: 1 1 auto;
+    min-height: 44px;
+    box-sizing: border-box;
+  }
+  #config-modal .mfa-code-input {
+    width: 100%;
+    max-width: 320px;
+    min-height: 52px;
+    font-size: 24px;
+    letter-spacing: 6px;
+  }
+  #config-modal .mfa-secret-code {
+    display: block;
+    word-break: break-all;
+    overflow-wrap: anywhere;
+  }
+  #config-modal .mfa-code-input-group {
+    flex-wrap: wrap;
+    justify-content: flex-start;
+  }
+  /* WebAuthn key rows: wrap instead of clipping the remove button. */
+  #config-modal .webauthn-key-item {
+    flex-wrap: wrap;
+    gap: 8px;
+  }
+  #config-modal .webauthn-key-label {
+    flex: 1 1 100%;
+    min-width: 0;
+  }
+  #config-modal .webauthn-key-item .config-btn-sm {
+    min-height: 44px;
+    min-width: 44px;
+  }
+  #config-modal .hidden-files-add-row {
+    flex-wrap: wrap;
+  }
+  #config-modal .hidden-files-add-row .config-input {
+    flex: 1 1 100%;
+    max-width: none;
+    min-width: 0;
+  }
+  #config-modal .config-diag-row {
+    flex-wrap: wrap;
+    gap: 4px 8px;
+  }
+  /* AI keys sticky footer: clear the mobile toolbar + home indicator. */
+  #config-modal .ai-keys-footer {
+    padding-bottom: calc(12px + env(safe-area-inset-bottom, 0));
+    bottom: 0;
  }
  /* Two-column grids → single column. */
  #config-modal .ai-default-grid,
@@ -4851,6 +5096,61 @@ body.resizing-v {
    min-width: 0;
    overflow-wrap: anywhere;
  }
+  /* Drawer: left slide-over above a dimming backdrop. Higher specificity
+     than the generic .help-nav rules (body .help-nav / .help-nav) that hide
+     the nav on mobile. JS toggles .config-toc-open on #config-modal. */
+  #config-modal #config-nav {
+    position: fixed;
+    top: 0;
+    left: 0;
+    bottom: 0;
+    width: min(320px, 88vw);
+    min-width: 0;
+    max-width: 88vw;
+    max-height: 100dvh;
+    height: 100%;
+    box-sizing: border-box;
+    z-index: 40;
+    background: var(--bg-secondary);
+    border-right: 1px solid var(--border);
+    border-bottom: none;
+    box-shadow: none;
+    backdrop-filter: none;
+    display: none;
+    overflow: hidden;
+    animation: config-toc-slide-in 200ms ease-out;
+  }
+  #config-modal.config-toc-open #config-nav {
+    display: flex;
+    max-height: 100dvh;
+  }
+  #config-modal .help-nav-list {
+    padding-bottom: calc(20px + env(safe-area-inset-bottom, 0));
+  }
+  /* Backdrop: covers the full modal, sits above the content (z-index auto)
+     and below the drawer (z-index 40). Clicks on the pseudo-element are
+     attributed to #config-modal, so the existing e.target === modal handler
+     closes the drawer first; it also blocks taps reaching the content. */
+  #config-modal.config-toc-open::before {
+    content: "";
+    position: absolute;
+    inset: 0;
+    z-index: 35;
+    background: rgba(0, 0, 0, 0.45);
+  }
+  @keyframes config-toc-slide-in {
+    from {
+      transform: translateX(-100%);
+    }
+    to {
+      transform: translateX(0);
+    }
+  }
+  /* Content area gets a bottom pad so the sticky Save row never covers the
+     last controls while the virtual keyboard is open. */
+  #config-modal #config-scroll .config-content {
+    padding-bottom: 80px;
+  }
 }

 /* --- Toast notifications --- */
@@ -5819,28 +6119,6 @@ body.resizing-v {
  font-size: 0.9rem;
 }

-.config-btn-add {
-  padding: 6px 12px;
-  background: var(--accent);
-  color: white;
-  border: none;
-  border-radius: 4px;
-  cursor: pointer;
-  font-size: 0.875rem;
-  font-weight: 500;
-  transition: all 150ms ease;
-  white-space: nowrap;
-}
-
-.config-btn-add:hover {
-  background: color-mix(in srgb, var(--accent) 85%, black);
-  transform: translateY(-1px);
-}
-
-.config-btn-add:active {
-  transform: translateY(0);
-}
-
 /* ---------------------------------------------------------------------------
   Utility — hidden class
   --------------------------------------------------------------------------- */
@@ -7907,7 +8185,6 @@ body.popup-mode .content-area {
 .webhook-events { color: var(--text-muted); font-size: 0.7rem; }
 .webhook-delete { background: none; border: none; color: var(--text-error); cursor: pointer; font-size: 1rem; padding: 2px 6px; }
 .config-add-row { display: flex; gap: 8px; margin-top: 8px; }
-.config-btn-add { padding: 6px 14px; background: var(--accent); color: #fff; border: none; border-radius: 6px; cursor: pointer; font-size: 0.8rem; }

 /* ── API/MCP tokens UI (#107) ── */
 .token-item {
@@ -8506,9 +8783,13 @@ body.desktop-mode .editor-container {

  .help-hamburger {
    display: inline-flex;
+    width: 44px;
+    height: 44px;
+    min-width: 44px;
  }

-  /* TOC: collapsible block at top (not slide-over) */
+  /* TOC drawer: collapsible block for the help modal; the config modal
+     (#114) overrides this with a fixed left drawer + backdrop. */
  .help-nav {
    position: static;
    width: 100%;
@@ -8617,13 +8898,10 @@ body.desktop-mode .editor-container {
  }

  /* Push content up to avoid toolbar overlap */
-  .main-layout {
+  .main-body {
    padding-bottom: 64px;
    padding-bottom: calc(64px + env(safe-area-inset-bottom, 0));
  }
-  .editor-modal.active ~ .main-layout {
-    padding-bottom: 0;
-  }

  /* Enable pull-to-refresh on mobile — unblock the overscroll chain */
  html {
@@ -10567,7 +10845,9 @@ body.desktop-mode .editor-container {
    bottom: calc(64px + env(safe-area-inset-bottom, 0px) + 12px);
  }
  body.np-active .content-area {
-    padding-bottom: calc(64px + env(safe-area-inset-bottom, 0px) + 76px);
+    /* Toolbar clearance is carried by .main-body (BUG-073): only the dock
+       zone (76px above the toolbar) remains here — same total as before. */
+    padding-bottom: 76px;
  }
 }

@@ -10644,6 +10924,205 @@ body.desktop-mode .editor-container {
  background: var(--surface);
 }

+/* ── XLSX Viewer ── */
+.xlsx-toolbar {
+  display: flex;
+  align-items: center;
+  gap: 10px;
+  margin-bottom: 8px;
+  flex-wrap: wrap;
+}
+.xlsx-toolbar-actions {
+  margin-left: auto;
+  display: flex;
+  gap: 8px;
+}
+.xlsx-tabs {
+  display: flex;
+  gap: 4px;
+  flex-wrap: wrap;
+}
+.xlsx-tab {
+  border: 1px solid var(--border);
+  background: var(--surface);
+  color: var(--text-secondary);
+  border-radius: 4px;
+  padding: 4px 10px;
+  font-size: 0.8rem;
+  cursor: pointer;
+}
+.xlsx-tab.active {
+  background: var(--accent, #4a90d9);
+  border-color: var(--accent, #4a90d9);
+  color: #fff;
+}
+.xlsx-table th.xlsx-corner,
+.xlsx-table th.xlsx-rownum {
+  background: var(--surface);
+  color: var(--text-secondary);
+  font-weight: 400;
+  text-align: right;
+  padding: 6px 8px;
+  border-bottom: 2px solid var(--border);
+  border-right: 1px solid var(--border-light, var(--border));
+  position: sticky;
+  left: 0;
+  /* #153 A8 — `top: auto` is load-bearing: `.csv-table th` pins EVERY `th`
+     at `top: 0`, so a row number left sticky on both axes piles up in the
+     top-left corner instead of tracking its own row. */
+  top: auto;
+  z-index: 2;
+}
+.xlsx-table th.xlsx-corner {
+  left: 0;
+  top: 0;
+  z-index: 4;
+}
+/* #153 A8 — the column headers stay visible while the sheet scrolls down.
+   Declared explicitly (and not inherited from `.csv-table th`) so the stacking
+   order is intentional: thead (3) < row numbers (2) < corner (4). */
+.xlsx-table thead th {
+  position: sticky;
+  top: 0;
+  z-index: 3;
+  background: var(--surface);
+}
+.xlsx-table td[contenteditable] {
+  cursor: text;
+  min-width: 40px;
+  white-space: pre-wrap;
+}
+.xlsx-table td[contenteditable]:focus {
+  outline: 2px solid var(--accent, #4a90d9);
+  outline-offset: -2px;
+}
+.xlsx-table td.xlsx-dirty {
+  background: rgba(255, 196, 0, 0.18);
+}
+
+/* #153 A12 — last result Excel computed, shown under a formula cell.
+   Discreet by design: the formula is what the user edits, the cached value is
+   context (stale until Excel recalculates). */
+.xlsx-cached {
+  display: block;
+  margin-top: 2px;
+  padding-left: 6px;
+  border-left: 2px solid var(--border, #d0d7de);
+  color: var(--text-muted);
+  font-size: 0.85em;
+  font-variant-numeric: tabular-nums;
+  white-space: nowrap;
+}
+
+/* #153 A1/A4 — lossy-save warning + formula toggle */
+.xlsx-warning {
+  display: flex;
+  align-items: flex-start;
+  gap: 8px;
+  padding: 8px 10px;
+  margin-bottom: 8px;
+  border: 1px solid var(--warning, #e0a800);
+  border-left-width: 3px;
+  border-radius: 4px;
+  background: var(--surface);
+  color: var(--text-secondary);
+  font-size: 0.82rem;
+  line-height: 1.45;
+}
+.xlsx-warning-icon {
+  width: 16px;
+  height: 16px;
+  flex: 0 0 auto;
+  margin-top: 1px;
+  color: var(--warning, #e0a800);
+}
+.xlsx-warning-body {
+  display: flex;
+  flex-direction: column;
+  gap: 3px;
+  min-width: 0;
+}
+.xlsx-warning-body strong {
+  color: var(--text-primary);
+  font-weight: 600;
+}
+.xlsx-warning-list {
+  display: flex;
+  flex-wrap: wrap;
+  gap: 4px;
+}
+.xlsx-warning-tag {
+  padding: 1px 6px;
+  border: 1px solid var(--border);
+  border-radius: 10px;
+  background: var(--bg-secondary);
+  color: var(--text-secondary);
+  font-size: 0.75rem;
+  white-space: nowrap;
+}
+.xlsx-warning-hint {
+  color: var(--text-secondary);
+  opacity: 0.85;
+}
+/* #153 A8 — "feuille tronquée" notice. Deliberately NOT the `.xlsx-warning`
+   look: that one is a data-loss alert, this one only says part of the sheet is
+   out of view. */
+.xlsx-truncated {
+  display: flex;
+  align-items: flex-start;
+  gap: 8px;
+  padding: 8px 10px;
+  margin-bottom: 8px;
+  border: 1px solid var(--border);
+  border-left: 3px solid var(--accent, #4a90d9);
+  border-radius: 4px;
+  background: var(--bg-secondary);
+  color: var(--text-secondary);
+  font-size: 0.82rem;
+  line-height: 1.45;
+}
+.xlsx-truncated-icon {
+  width: 16px;
+  height: 16px;
+  flex: 0 0 auto;
+  margin-top: 1px;
+  color: var(--accent, #4a90d9);
+}
+
+/* #153 A9bis — “charger la suite” footnote under a truncated sheet. Also the
+   scroll sentinel target: clickable whole, disabled look once the sheet is
+   fully loaded. */
+.xlsx-load-more {
+  display: block;
+  margin: 6px 0 10px;
+  padding: 6px 12px;
+  border: 1px dashed var(--border);
+  border-radius: 4px;
+  background: var(--bg-secondary);
+  color: var(--text-secondary);
+  font-size: 0.82rem;
+  text-align: center;
+  cursor: pointer;
+  user-select: none;
+}
+.xlsx-load-more:hover {
+  border-color: var(--accent, #4a90d9);
+  color: var(--text-primary);
+}
+.xlsx-load-more.done {
+  display: none;
+}
+
+.xlsx-formula-toggle {
+  font-family: 'JetBrains Mono', 'Fira Code', 'Consolas', monospace;
+  font-weight: 600;
+}
+.xlsx-formula-toggle.active {
+  background: var(--accent, #4a90d9);
+  border-color: var(--accent, #4a90d9);
+  color: #fff;
+}
+
 /* ── JSON Viewer ── */
 .json-viewer {
  font-family: 'JetBrains Mono', 'Fira Code', 'Consolas', monospace;
@@ -10764,7 +11243,6 @@ body.desktop-mode .editor-container {
 }
 .mfa-link-btn:hover { opacity: 0.8; }
 .mfa-error { color: #e74c3c; font-size: 13px; margin-top: 8px; }
-.mfa-recovery-input { width: 200px; font-size: 20px; letter-spacing: 4px; }

 /* MFA Settings (Security tab) */
 .mfa-status-section { padding: 16px 0; }
@@ -10963,6 +11441,9 @@ body.desktop-mode .editor-container {
  justify-content: center; flex-shrink: 0; }
 .bookslm-input-area button:hover { opacity: 0.9; }
 .bookslm-input-area button:disabled { opacity: 0.5; cursor: not-allowed; }
+/* BUG-077 — the send button doubles as a Stop control while streaming. */
+.bookslm-input-area button.bookslm-btn-send.is-stopping { background: var(--danger); }
+.bookslm-input-area button.bookslm-btn-send.is-stopping:hover { opacity: 1; filter: brightness(1.08); }
 .bookslm-input-hint { padding: 0 16px 10px; font-size: 10px; color: var(--text-secondary);
  text-align: right; border-top: none; }
 /* Action cards proposed by the General assistant (file/dir creation). */
@@ -10991,6 +11472,9 @@ body.desktop-mode .editor-container {
 .bookslm-tool-trace > summary:hover { color: var(--text-primary); }
 .bookslm-tool-trace[open] > summary { margin-bottom: 4px; }
 .bookslm-steps-dots { color: var(--accent); }
+/* BUG-074 — preview of the first action next to the step counter. */
+.bookslm-steps-title { color: var(--text-secondary); max-width: 45vw; overflow: hidden;
+  text-overflow: ellipsis; white-space: nowrap; }
 .bookslm-steps-body { display: flex; flex-direction: column; gap: 3px; }
 .bookslm-chevron { font-size: 8px; line-height: 1; opacity: 0.7; }
 .bookslm-chevron::before { content: '▶'; }
@@ -11026,6 +11510,15 @@ details[open] > summary .bookslm-chevron::before { content: '▼'; }
 /* Mutation confirmation card (two-step propose/apply). */
 .bookslm-confirm { flex-wrap: wrap; }
 .bookslm-confirm-diff { flex-basis: 100%; width: 100%; margin-top: 4px; }
+/* BUG-075 — one row per action when the run batches several mutations. */
+.bookslm-confirm-actions { flex-basis: 100%; width: 100%; display: flex;
+  flex-direction: column; gap: 4px; margin-top: 4px; }
+.bookslm-confirm-action > summary { display: inline-flex; align-items: center; gap: 5px;
+  font-size: 12px; color: var(--text-secondary); cursor: pointer; list-style: none;
+  word-break: break-word; }
+.bookslm-confirm-action > summary::-webkit-details-marker { display: none; }
+.bookslm-confirm-action > summary:hover { color: var(--text-primary); }
+.bookslm-confirm-action[open] > summary { color: var(--text-primary); }
 .bookslm-diff-title { font-size: 11px; color: var(--text-secondary); margin-bottom: 4px; }
 .bookslm-diff { background: rgba(0,0,0,0.25); border-radius: 6px; padding: 8px;
  overflow-x: auto; font-size: 12px; line-height: 1.4; max-height: 240px; margin: 0; }
@@ -11600,7 +12093,12 @@ body.reading-mode .content-area {
  margin: 0 auto;
  padding: clamp(20px, 6vw, 48px) clamp(18px, 6vw, 40px) 96px;
 }
+/* The bottom toolbar is hidden in reading mode: no clearance needed. */
+body.reading-mode .main-body {
+  padding-bottom: 0;
+}
 body.reading-mode .breadcrumb,
+body.reading-mode .file-toolbar,
 body.reading-mode .file-actions {
  display: none;
 }
@@ -11,7 +11,7 @@
 * cache or Cloudflare does NOT clear the Service Worker Cache Storage, which is
 * a separate store. Bumping SW_VERSION invalidates it on every release.
 */
-const SW_VERSION = 'v24';
+const SW_VERSION = 'v27';
 const CODE_CACHE = `obsigate-code-${SW_VERSION}`;
 const RUNTIME_CACHE = `obsigate-runtime-${SW_VERSION}`;
 const API_CACHE = `obsigate-api-${SW_VERSION}`;
@@ -1,6 +1,6 @@
 {
  "name": "obsigate",
-  "version": "2.22.0",
+  "version": "2.33.0",
  "description": "**Porte d'entrée web ultra-léger pour vos vaults Obsidian** — Accédez, naviguez et recherchez dans toutes vos notes Obsidian depuis n'importe quel appareil via une interface web moderne et responsive.",
  "main": "patch.js",
  "directories": {
@@ -1,4 +1,9 @@
-# Plan: Incremental InvertedIndex for 40k+ files
+# Incremental InvertedIndex for 40k+ files — livré
+
+> **Statut : LIVRÉ (BUG-033, v2.3.0).** Ce fichier a servi de plan
+> d'exécution ; il est conservé comme **trace de conception**. Le code réel
+> a divergé sur plusieurs points (voir [État réel](#état-réel-corrigé-au-2026-09-27))
+> — ne pas lire les extraits de code ci-dessous comme du code actuel.

 ## Problem Summary

@@ -17,6 +22,11 @@ Then hook these into `_add_file_to_structures` and `_remove_file_from_structures

 Remove the `is_stale()` / `rebuild()` / cooldown mechanism entirely. The inverted index is always current.

+> ⚠️ **Nuance retenue à l'implémentation** : un unique `rebuild()` reste nécessaire au
+> démarrage (le hook est inerte tant que l'index n'est pas prêt) et au reindex manuel
+> d'une vault. Ce qui disparaît, c'est la *staleness* : plus de compteur de génération,
+> plus de cooldown, plus de rebuild paresseux.
+
 ## Dependency Architecture

 **Current import chain:**
@@ -346,3 +356,61 @@ This hack was only needed to reduce the number of inverted index rebuilds. With
 4. **Sorted tokens performance:** `bisect.insort` and `list.pop(idx)` are O(V) worst case for large V. For 40k files, the vocabulary size V is typically 50k-200k tokens. O(V) for a single insertion is ~0.001ms, acceptable. The rebuild() call at startup handles the initial bulk.

 5. **tag_norm_map / tag_prefix_index growth:** These grow monotonically (never shrink on incremental remove). With 40k files and thousands of tags, this is a few thousand entries — negligible. A manual "Réindexer" button triggers a full `rebuild()` to clean up.
+
+---
+
+## État réel (corrigé le 2026-09-27)
+
+Le plan ci-dessus a servi de brouillon : **le code livré en est différent sur
+quatre points**. Relevé fait sur `backend/search.py`, `backend/indexer.py` et
+`backend/main.py`, pas de mémoire.
+
+| Point prévu | État réel |
+|---|---|
+| Étapes 1-2 : hook + `add_document()` / `remove_document()` | ✅ livré tel que prévu |
+| Étapes 4-5 : `rebuild()` initial via `init_inverted_index()` appelé depuis la lifespan | ✅ livré (`backend/main.py:297`, dans l'exécuteur de recherche) |
+| Étape 6 : retirer `is_stale()` + `_last_rebuild` / `_rebuild_cooldown` / `_source_generation` | ✅ **déjà fait** avant cette relecture — aucun de ces symboles ne subsiste |
+| Étape 7 : retirer le hack de coalescence `_index_generation` dans `_on_vault_change` | ✅ **déjà fait** — `_on_vault_change` n'existe plus |
+| `get_inverted_index()` simplifié | ✅ mais **sans le fallback `_needs_rebuild`** prévu par le plan |
+
+### Écarts assumés
+
+1. **`is_stale()` a survécu sous un autre nom.** L'étape 6 est faite, mais la
+   méthode a été conservée car elle répond à une autre question : *l'index
+   initial est-il construit ?* Elle ne mesure plus aucune staleness (le compteur
+   de génération et le cooldown ont disparu) et le nom était trompeur. Elle est
+   donc renommée `is_ready()` — cohérent avec le `is_ready()` déjà exposé par
+   `SemanticIndex` (`backend/semantic_search.py`). L'alias `is_stale()` de
+   `SemanticIndex`, sans aucun appelant, est supprimé.
+
+   Impact : le champ de `/api/diagnostics` passe de `is_stale` à `is_ready`
+   (libellé « Index prêt » côté `frontend/js/config.js`).
+
+2. **Pas de repli `_needs_rebuild`.** Le plan prévoyait qu'un échec
+   d'incrémentation marque l'index pour reconstruction. L'implémentation
+   retenue se contente de logger un warning et de continuer à servir l'index.
+   Choix assumé : un échec d'incrémentation est exceptionnel, et reconstruire
+   silencieusement serait plus coûteux que l'état dégradé. **Si ce compromis
+   devient critiquique, c'est le point à rouvrir.**
+
+3. **`_ready` remplace `doc_count == 0`.** Le plan prévoyait de sauter le hook
+   « index vide » ; le drapeau explicite `_ready` est plus sûr (un vault
+   réellement vide serait sinon pris pour un index non construit).
+
+4. **`rebuild()` reste nécessaire** au démarrage et au reindex manuel d'une
+   vault. Le plan parlait de le supprimer de `get_inverted_index()`, ce qui est
+   fait, mais la méthode elle-même est conservée.
+
+### Bug trouvé pendant cette relecture (corrigé ici)
+
+`remove_vault_from_index()` (`backend/indexer.py`) ne notifiait pas le hook.
+Conséquence mesurée : après suppression d'une vault, ses 8 documents test
+restaient dans l'index inversé — `postings`, `doc_info`, `doc_vault`,
+`vault_docs` — et continuaient de correspondre aux recherches pour une vault
+inexistante. Seul un reindex manuel les effaçait.
+
+Le correctif déclenche `_on_index_change('remove', …)` pour chaque fichier de
+la vault, et `_remove_doc_internals()` supprime désormais la clé `vault_docs`
+quand son set devient vide (c'est un `defaultdict` : une simple lecture la
+ré créait). Test de non-régression :
+`TestVaultRemovalPurgesInvertedIndex` (contre-preuve : échoue sans le patch).
@@ -9,6 +9,15 @@ export default defineConfig({
  reporter: process.env.CI ? 'github' : 'list',
  timeout: 60000,
  expect: { timeout: 10000 },
+  // BUG-080 : la suite (~120 tests, workers: 1, ~10-15 s/test sur un poste
+  // chargé) ne doit jamais pendre toute la nuit. Au-delà du timeout global,
+  // Playwright abandonne avec un échec explicite au lieu de bloquer.
+  // Surchargable : E2E_GLOBAL_TIMEOUT_MS.
+  globalTimeout: Number(
+    process.env.E2E_GLOBAL_TIMEOUT_MS ??
+      (process.env.CI ? 30 * 60 * 1000 : 25 * 60 * 1000),
+  ),
+  reportSlowTests: process.env.CI ? null : { max: 5, threshold: 30000 },

  use: {
    baseURL: process.env.BASE_URL || 'http://localhost:2029',
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
bruno	011ec84f23	feat: outils IA de lecture et d'edition des classeurs existants #153 A6 : list_xlsx_sheets (noms + dimensions + truncated), xlsx_to_markdown (table bornee 100x20 pour le contexte LLM), update_xlsx_cells et append_xlsx_rows (WRITE + confirmation, via edit_xlsx_cells : verrou, ecriture atomique, neutralisation des formules, 409 lossy). Les lignes ajoutees sont coercees comme dans le viewer (A10). Labels de step ai.step.xlsx_* FR/EN ; update_xlsx_cells/append_xlsx_rows branches sur MUTATING_TOOLS et FILE_WRITE_TOOLS (refresh viewer via obsigate:file-written). Tests : test_spreadsheet_tools.py 17 (risque, confirmation, persistance, garde formule heritee, coercion) + contre-preuve (cells vide -> 2 echecs). ruff 0, mypy 0, i18n parity, validate-imports. 🤖 Generated with Codebuff Co-Authored-By: Codebuff <[email protected]>	2026-09-28 12:05:16 -04:00
bruno	472ea9d309	feat: chargement à la demande des lignes cachées des tableurs #153 CI / lint (push) Successful in 2m24s Details CI / security (push) Failing after 1m39s Details CI / test (push) Successful in 4m0s Details CI / build (push) Successful in 1m37s Details CI / e2e (push) Successful in 15m34s Details A9bis : sous une feuille tronquée, un pied de page « N lignes affichées sur M · Charger la suite » fetch la fenêtre suivante (limit=500) au clic ou à l'approche du bas du tableau (sentinelle de défilement, marge 120px). Les lignes ajoutées passent par le même pipeline d'édition que le rendu initial (setupCell factorisé) : éditables et sauvegardables immédiatement. Fetch échoué → bouton restauré (retry) + toast ; feuille complète → pied de page masqué (class done). Vérifié : xlsx-viewer.test.mjs 19/19 (5 nouveaux, contre-preuve wireLazyRows désactivé → 5 échecs), E2E 8/8 (A520 visible et éditable après clic), validate-imports 40 modules, unit.test.mjs 12/12, i18n parity. 🤖 Generated with Codebuff Co-Authored-By: Codebuff <[email protected]>	2026-09-28 11:52:42 -04:00
bruno	6ba04c4381	feat: troncature annoncee et lecture par fenetres des tableurs #153 CI / lint (push) Successful in 2m29s Details CI / security (push) Failing after 1m39s Details CI / test (push) Successful in 4m1s Details CI / build (push) Successful in 1m37s Details CI / e2e (push) Successful in 14m47s Details - BUG-090 (#153 A8) : render_sheets() expose total_rows/total_cols, max_rows/max_cols et truncated ; la visionneuse affiche un bandeau « Feuille tronquée » (i18n FR/EN) au lieu de couper en silence, et la ligne d'en-têtes devient sticky (top:auto sur les numéros de ligne). - #153 A9 : GET /api/file/{vault}/xlsx/sheet?sheet&offset&limit sert une fenêtre de 1 à 1000 lignes avec les vraies coordonnées A1, has_more de pagination et valeurs calculées A12 ; 404 feuille inconnue, 415 non-xlsx. - Tests : TestXlsxTruncationNotice (4) + TestXlsxSheetWindow (11) avec contre-preuves, xlsx-viewer.test.mjs 14/14, E2E 7/7 (fixture sample-xlsx-large.xlsx 520 lignes), suite 1417 passed / 6 skipped, ruff/mypy 0, i18n parity. 🤖 Generated with Codebuff Co-Authored-By: Codebuff <[email protected]>	2026-09-28 10:38:24 -04:00
bruno	06f8e63d06	feat: tableurs indexables, saisie typée et valeurs calculées #153 Les tableurs étaient invisibles à la recherche, chaque saisie devenait du texte, et une formule n'affichait que sa formule. - A5 : extract_indexable_text() indexe les noms de feuilles et les 20 premières lignes (plafond 5 k caractères) pour le TF-IDF et la recherche sémantique. Un mot tapé dans une cellule rend le fichier trouvable ; un classeur chiffré s'indexe par son seul nom. - A10 : la saisie est typée comme dans Excel — booléens (TRUE/FAUX/OUI/NON) et dates FR JJ/MM/AAAA en ordre jour-first, donc 01/02/2026 est le 1er février. Une saisie ressemblant à une formule n'est jamais convertie (BUG-088 préservé). - A12 : la valeur calculée par Excel s'affiche sous la formule, via une 2e lecture data_only=True faite seulement si l'archive contient un <v>. Info-bulle traduite FR/EN, aucun texte d'interface côté backend. - BUG-089 : un reindex manuel ne reconstruisait pas l'index inversé, et backend/search.py lisait l'index via un import par valeur — après un rechargement du module, la recherche écrivait dans un dict périmé. Contre-preuves vérifiées pour A5, A10 et A12 (neutralisation de chaque fonction → échec des tests concernés). Tests : 1402 pytest, 10 JSDOM, 4 E2E, suite E2E complète verte, ruff/mypy 0. 🤖 Generated with Codebuff Co-Authored-By: Codebuff <[email protected]>	2026-09-27 22:43:24 -04:00
bruno	31d4616baf	feat: garde-fous d'écriture des classeurs Excel #153 (P0) L'édition d'un .xlsx pouvait détruire une partie du classeur, le concurrencer en silence, ou diffuser une injection de formule. - BUG-085 : inspect_workbook() détecte ce qu'un round-trip openpyxl perd (valeurs calculées en cache, slicers, contrôles, connexions, custom XML, signature, commentaires enrichis, macros) → xlsx_lossy_features exposé en lecture, bandeau FR/EN, et 409 xlsx_lossy_content sans `force` (confirmation explicite puis reprise). Périmètre réel revalidé : graphiques, images et TCD survivent au round-trip. - BUG-086 : écriture atomique (fichier .tmp + os.replace) : un plantage ne peut plus tronquer le classeur, le backup reste intact. - BUG-087 : verrou par fichier autour du read-modify-write (timeout 15 s, 409 conflict) ; endpoint xlsx/save devenu synchrone pour que l'attente s'exécute dans le threadpool. - BUG-088 : une saisie en '=' ou '@' est stockée en texte, sauf opt-in `allow_formula` ou le bouton f(x) de la visionneuse. Le handler ServiceError expose désormais code + details, que api() propage. - BUG-084 : la suppression d'une vault purge enfin l'index inversé (documents fantômes qui continuaient de matcher) et is_stale() devient is_ready(), le nom étant trompeur (la staleness n'existe plus). Tests : 1390 pytest, 10 JSDOM (xlsx-viewer.test.mjs, branché au CI), 3 E2E Playwright, suite E2E complète verte, ruff/mypy 0. 🤖 Generated with Codebuff Co-Authored-By: Codebuff <[email protected]>	2026-09-27 20:39:12 -04:00
bruno	4c4b1222d5	chore: ignore les diagrammes E2E générés (excalidraw) CI / lint (push) Successful in 2m23s Details CI / security (push) Failing after 1m48s Details CI / test (push) Successful in 3m49s Details CI / build (push) Successful in 2m28s Details CI / e2e (push) Successful in 14m5s Details	2026-09-27 14:46:09 -04:00
bruno	a3973b981c	securite: #87 T6-T8 fin dette — deps qualifiées, semgrep, Secure auto, CORS	2026-09-27 12:43:31 -04:00
bruno	b6e2029770	fix: E2E node-direct, timeouts réalistes 25/30 min (complément BUG-080) CI / lint (push) Successful in 2m22s Details CI / security (push) Successful in 1m37s Details CI / test (push) Successful in 4m48s Details CI / build (push) Successful in 1m33s Details CI / e2e (push) Successful in 14m3s Details	2026-09-27 11:28:35 -04:00
bruno	6b878caff3	securite: #87 T5c script-src sans unsafe-inline (nonces T5b) CI / lint (push) Successful in 2m18s Details CI / security (push) Successful in 1m37s Details CI / test (push) Successful in 4m32s Details CI / build (push) Successful in 1m34s Details CI / e2e (push) Successful in 14m22s Details	2026-09-27 10:36:00 -04:00
bruno	e9b7a317c1	fix: mfa/status 200 auth désactivée (garde anonymous) BUG-081 + clôture BUG-080/082/083	2026-09-27 10:35:33 -04:00
bruno	14b8032635	fix: CI lint — config-ai-keys.test.mjs rejoint l'étape JSDOM (complément BUG-082) CI / security (push) Successful in 1m37s Details CI / lint (push) Successful in 2m18s Details CI / test (push) Successful in 4m8s Details CI / build (push) Successful in 1m38s Details CI / e2e (push) Successful in 14m28s Details	2026-09-27 09:44:18 -04:00
bruno	7dfe26c83d	fix: CI security — echo pip-audit sans dièse (runner Act) BUG-083 CI / lint (push) Failing after 1m50s Details CI / test (push) Skipped Details CI / build (push) Skipped Details CI / e2e (push) Skipped Details CI / security (push) Successful in 1m37s Details	2026-09-27 09:37:34 -04:00
bruno	24229316c7	fix: CI lint — upload.test.mjs rejoint l'étape JSDOM (jsdom) BUG-082 CI / lint (push) Failing after 1m30s Details CI / test (push) Skipped Details CI / build (push) Skipped Details CI / e2e (push) Skipped Details CI / security (push) Failing after 1m33s Details	2026-09-27 09:23:51 -04:00
bruno	7d70e0fb75	fix: harnais E2E local anti-blocage BUG-080 CI / lint (push) Failing after 1m51s Details CI / test (push) Skipped Details CI / build (push) Skipped Details CI / e2e (push) Skipped Details CI / security (push) Failing after 1m33s Details	2026-09-27 09:08:58 -04:00
bruno	d70ecd0968	securite: #87 T5b nonces CSP prets pour bascule (sans changement)	2026-09-26 22:44:12 -04:00
bruno	36a4030c09	securite: #87 T5a supprime 16 handlers inline au profit de listeners (CSP inchangee)	2026-09-26 22:32:01 -04:00
bruno	330462e7a5	docs: #87 T4 consigne resultats E2E locaux (108/108 desktop, 10/10 mobiles cibles)	2026-09-26 22:22:31 -04:00
bruno	922dfa2e79	test: #87 T4 E2E XSS partage et lecteur + script serveur E2E avec progression	2026-09-26 21:46:16 -04:00
bruno	34fce932cb	securite: #87 T3 cookies Secure centralises + CORS atteste (defaut inchange)	2026-09-26 18:37:19 -04:00
bruno	18b1e13f34	test: #87 T2 durcissement concurrence users.json et budget temps regex	2026-09-26 18:34:06 -04:00
bruno	7bee4a237d	ci: #87 T1 bandit et npm audit bloquants, 5 suites frontend au CI	2026-09-26 18:30:04 -04:00
bruno	d6cca2b1af	feat: #85 T10 persistance etat (verrous stores, ratelimit SQLite) et cloture refonte	2026-09-26 18:10:10 -04:00
bruno	58312e64da	refactor: #85 T9 extrait realtime et render vers modules dedies (comportement inchange)	2026-09-26 16:59:25 -04:00
bruno	3b0927a8c9	refactor: #85 T8 extrait vaults-history-conflicts vers backend/routers (comportement inchange)	2026-09-26 14:46:55 -04:00
bruno	6e527c371d	refactor: #85 T7 extrait le domaine config vers backend/routers (comportement inchange)	2026-09-26 14:33:40 -04:00
bruno	6cccdc1f34	refactor: #85 T6c extrait media-pdf-export vers backend/routers (comportement inchange)	2026-09-26 14:06:52 -04:00
bruno	0abc17e9f2	refactor: #85 T6b extrait mutations fichiers-dossiers vers backend/routers (comportement inchange)	2026-09-26 13:51:36 -04:00
bruno	b83d8dacdf	refactor: #85 T6a extrait lecture fichiers vers backend/routers (comportement inchange)	2026-09-26 13:43:53 -04:00
bruno	dadc055429	refactor: #85 T5 extrait le domaine search vers backend/routers + executor partage (comportement inchange)	2026-09-26 13:14:31 -04:00
bruno	750114a923	refactor: #85 T4 extrait le domaine backups vers backend/routers + sse partage (comportement inchange)	2026-09-26 13:06:45 -04:00
bruno	83a81da319	refactor: #85 T3 extrait le domaine sharing vers backend/routers (comportement inchange)	2026-09-26 12:51:06 -04:00
bruno	3eb0256127	refactor: #85 T2 extrait le CRUD webhooks vers backend/routers (comportement inchange)	2026-09-26 12:43:57 -04:00
bruno	9d8b3cc854	refactor: #85 T1 extrait le domaine health vers backend/routers (comportement inchange)	2026-09-26 12:36:49 -04:00
bruno	0ab402aa73	docs: roadmap priorise dette et securite #85 #87 (#77 non signe, #73 reporte) CI / lint (push) Successful in 2m13s Details CI / security (push) Successful in 1m34s Details CI / test (push) Successful in 4m9s Details CI / build (push) Successful in 1m31s Details CI / e2e (push) Successful in 14m17s Details	2026-09-26 11:55:32 -04:00
bruno	c36c299466	docs: #152 — aligne changelog et roadmap sur la version livrée 2.27.0 Fusionne la section 2.26.0 (jamais publiée) dans 2.27.0 et corrige la ligne index de la roadmap ; suppression du tag local v2.26.0.	2026-09-25 20:32:14 -04:00
bruno	8611416670	feat: #152 viewer XLSX — affichage multi-feuilles, édition des cellules et téléchargement des .xlsx - backend/xlsx_reader.py : rend openpyxl en tableaux HTML (plafond 500x40 par feuille) - PUT /api/file/{vault}/xlsx/save : service edit_xlsx_cells (backup avant écriture, refs A1 validées) - frontend : renderXlsxViewer (onglets, contenteditable, sauvegarde par feuille) - docs : CHANGELOG [Unreleased], Roadmap index #152, archive, README FR/EN, exemple OpenAPI	2026-09-25 20:30:45 -04:00
bruno	e20fd6bf97	fix: /api/diagnostics 500 « dictionary changed size during iteration » — snapshot des dicts avant itération BUG-079 CI / lint (push) Successful in 2m3s Details CI / security (push) Successful in 1m27s Details CI / test (push) Successful in 4m22s Details CI / build (push) Successful in 1m22s Details CI / e2e (push) Successful in 13m36s Details	2026-09-24 17:13:40 -04:00
bruno	943005328c	feat: barre d'outils de lecture epinglee, coloration syntaxique des fichiers de code et avatars predefinis (#115 , #117 , BUG-078) CI / lint (push) Successful in 2m5s Details CI / security (push) Successful in 1m27s Details CI / test (push) Successful in 4m10s Details CI / build (push) Successful in 1m23s Details CI / e2e (push) Successful in 13m55s Details	2026-09-24 16:21:45 -04:00
bruno	e1842043d8	feat: assistant IA — approbation groupee des actions, bloc d'etapes, refresh UI et bouton Stop (BUG-074, BUG-075, BUG-076, BUG-077) CI / lint (push) Successful in 2m1s Details CI / security (push) Successful in 1m25s Details CI / test (push) Successful in 4m19s Details CI / build (push) Successful in 1m26s Details CI / e2e (push) Successful in 13m38s Details	2026-09-24 10:05:32 -04:00
bruno	9fb094f505	feat: refonte mobile de la section Configurations #114 CI / lint (push) Successful in 2m1s Details CI / security (push) Successful in 1m26s Details CI / test (push) Successful in 4m50s Details CI / build (push) Successful in 1m22s Details CI / e2e (push) Successful in 15m34s Details	2026-09-24 08:43:18 -04:00
bruno	d142049216	fix: degagement mobile sous la barre de navigation du bas (clearance retargetee .main-body) BUG-073 CI / lint (push) Successful in 1m59s Details CI / security (push) Successful in 1m26s Details CI / test (push) Successful in 3m41s Details CI / build (push) Successful in 1m21s Details CI / e2e (push) Successful in 14m0s Details	2026-09-23 23:36:35 -04:00
@@ -1 +1 @@
 .22.0
 .33.0