nicolasvd · nicolasvd · Sep 25, 2026 · Sep 25, 2026
diff --git a/.agent/hooks/pre-invocation-anchor.sh b/.agent/hooks/pre-invocation-anchor.sh
@@ -10,7 +10,7 @@ fi
 
 if [ -n "$BRANCH" ] && [ "$BRANCH" != "$DEF" ] && [ "$BRANCH" != "main" ] && [ "$BRANCH" != "master" ]; then
     cat <<EOF
-{"injectSteps":[{"ephemeralMessage":"[SYSTEM CONTEXT: Branche active: '$BRANCH' | État: Code / PR Review Loop. Applique toute correction ou retour de test directement sur cette branche. Ne PAS créer de nouvelle issue ni de nouvelle branche.]"}]}
+{"injectSteps":[{"ephemeralMessage":"[SYSTEM CONTEXT: Active branch: '$BRANCH' | State: Code / PR Review Loop. Apply all fixes or test feedback directly on this branch. Do NOT create a new issue or branch.]"}]}
 EOF
 else
     echo '{"injectSteps":[]}'

diff --git a/.github/ISSUE_TEMPLATE/bug_report.yml b/.github/ISSUE_TEMPLATE/bug_report.yml
@@ -1,70 +1,70 @@
-name: 🐛 Bug Report / Rapport d'Anomalie
-description: Signaler un dysfonctionnement, une régression ou un problème de synchronisation
+name: 🐛 Bug Report
+description: Report a malfunction, regression, or synchronization issue
 title: "fix(<scope>): <short descriptive bug title>"
 labels: ["type:bug"]
 body:
   - type: markdown
     attributes:
       value: |
-        ### 🔍 Signalement d'anomalie Agentic Android Kernel
-        Merci de détailler le comportement inattendu pour faciliter le diagnostic et la correction.
+        ### 🔍 Agentic Android Kernel Bug Report
+        Please detail the unexpected behavior to facilitate diagnosis and resolution.
   - type: dropdown
     id: scope
     attributes:
-      label: Périmètre / Domaine affecté
-      description: Choisissez le domaine principal où le bug se produit.
+      label: Scope / Affected Domain
+      description: Choose the primary domain where the bug occurs.
       options:
-        - tasks (Création, édition, suppression, complétion de charge)
-        - matrix (Affichage ou interaction matrice Eisenhower)
-        - duo (Synchronisation en temps réel, association couple)
-        - analytics (Calcul de jauge, tendances, graphiques)
-        - auth (Connexion Google, sessions)
-        - sync (Firestore, Room, hors-ligne)
-        - ui (Affichage, glitch visuel, animation)
-        - ai (Classification Gemini, timeout API)
+        - tasks (Creation, editing, deletion, mental load completion)
+        - matrix (Eisenhower matrix display or interaction)
+        - duo (Real-time synchronization, duo pairing)
+        - analytics (Gauge calculation, trends, charts)
+        - auth (Google Sign-In, sessions)
+        - sync (Firestore, Room, offline)
+        - ui (Display, visual glitch, animation)
+        - ai (Gemini classification, API timeout)
     validations:
       required: true
   - type: textarea
     id: bug-description
     attributes:
-      label: Description du Bug
-      description: Description claire et concise de l'anomalie rencontrée.
-      placeholder: "Exemple : Lors du changement d'onglet vers Duo, l'état de chargement reste figé..."
+      label: Bug Description
+      description: Clear and concise description of the issue encountered.
+      placeholder: "Example: When switching to the Duo tab, the loading state remains stuck..."
     validations:
       required: true
   - type: textarea
     id: steps-to-reproduce
     attributes:
-      label: Étapes pour Reproduire
-      description: Quelles sont les actions exactes menant au problème ?
+      label: Steps to Reproduce
+      description: What are the exact steps leading to the problem?
       placeholder: |
-        1. Ouvrir l'application
-        2. Aller sur l'onglet 'Matrice'
-        3. Glisser une tâche vers le quadrant 'Déléguer'
-        4. Observer l'erreur...
+        1. Open the application
+        2. Go to the 'Matrix' tab
+        3. Drag a task to the 'Delegate' quadrant
+        4. Observe the error...
     validations:
       required: true
   - type: textarea
     id: expected-vs-actual
     attributes:
-      label: Comportement Attendu vs Comportement Observé
-      description: Qu'aurait-il dû se passer et que s'est-il passé ?
-      placeholder: "Attendu : La tâche change de quadrant avec une animation fluide.\nObservé : Crash ou tâche inchangée."
+      label: Expected vs. Actual Behavior
+      description: What should have happened, and what actually occurred?
+      placeholder: "Expected: Task moves to target quadrant with smooth animation.\nActual: Crash or task remains unchanged."
     validations:
       required: true
   - type: input
     id: environment
     attributes:
-      label: Environnement de Test
-      description: Appareil, version d'Android, version de l'application.
-      placeholder: "Pixel 8 Pro - Android 14 / Émulateur API 34 - v0.1.0"
+      label: Test Environment
+      description: Device model, Android OS version, application version.
+      placeholder: "Pixel 8 Pro - Android 14 / Emulator API 34 - v0.1.0"
     validations:
       required: false
   - type: textarea
     id: logs-and-stacktrace
     attributes:
-      label: Logs / Stacktrace (Optionnel)
-      description: Collez ici les logs Android Studio / Logcat pertinents.
+      label: Logs / Stacktrace (Optional)
+      description: Paste relevant Android Studio / Logcat logs here.
       placeholder: "Logcat / Stacktrace..."
     validations:
       required: false
diff --git a/.github/ISSUE_TEMPLATE/feature_idea.yml b/.github/ISSUE_TEMPLATE/feature_idea.yml
@@ -1,71 +1,71 @@
-name: 💡 Feature Proposal / Idée de Fonctionnalité
-description: Proposer une nouvelle fonctionnalité ou amélioration pour Agentic Android Kernel
+name: 💡 Feature Proposal
+description: Propose a new feature or improvement for Agentic Android Kernel
 title: "feat(<scope>): <short descriptive title>"
 labels: ["type:feature"]
 body:
   - type: markdown
     attributes:
       value: |
-        ### 🧠 Proposition d'évolution Agentic Android Kernel
-        Merci de contribuer à la réduction de la charge mentale ! Veuillez remplir ce formulaire pour structurer la fonctionnalité selon la démarche Spec-Driven (4 Piliers).
+        ### 🧠 Feature Proposal for Agentic Android Kernel
+        Thank you for contributing to mental load reduction! Please fill out this form to structure your feature following the Spec-Driven (4 Pillars) framework.
   - type: dropdown
     id: scope
     attributes:
-      label: Périmètre / Domaine fonctionnel
-      description: Choisissez le domaine principal concerné par cette fonctionnalité.
+      label: Scope / Functional Domain
+      description: Choose the primary domain affected by this feature.
       options:
-        - tasks (Gestion des charges, récurrence, cycles)
-        - matrix (Matrice Eisenhower, priorisation 4 quadrants)
-        - duo (Mode Couple, rotation, synchronisation partagée)
-        - analytics (Graphiques de tendances, équilibre mental, jauges)
-        - core (Base de données locale Room, synchronisation Firestore, Auth)
-        - ai (Gemini classification automatique, suggestions intelligentes)
-        - ui (Composants visuels, animations sereines, Thème)
+        - tasks (Mental load management, recurrence, cycles)
+        - matrix (Eisenhower matrix, 4-quadrant prioritization)
+        - duo (Couple mode, task rotation, shared synchronization)
+        - analytics (Trend charts, mental balance, gauges)
+        - core (Local Room database, Firestore synchronization, Auth)
+        - ai (Gemini automated classification, smart suggestions)
+        - ui (Visual components, serene animations, Theme)
     validations:
       required: true
   - type: textarea
     id: problem-context
     attributes:
-      label: Contexte & Problématique Charge Mentale
-      description: Quel est le problème réel ou la friction vécue par l'utilisateur ou le duo ? En quoi cela alourdit-il la charge mentale ?
-      placeholder: "Exemple : Les tâches ménagères récurrentes sont souvent réassignées manuellement, provoquant des frictions et une asymétrie de charge..."
+      label: Context & Mental Load Problem Statement
+      description: What is the real problem or friction experienced by the user or duo? How does it increase mental load?
+      placeholder: "Example: Recurring household tasks are often manually reassigned, causing friction and asymmetric load..."
     validations:
       required: true
   - type: textarea
     id: proposed-solution
     attributes:
-      label: Solution Proposée & Expérience Utilisateur (UX)
-      description: Décrivez précisément la solution souhaitée, les interactions utilisateur et le résultat attendu.
-      placeholder: "Exemple : Introduire une règle de rotation alternée automatique à chaque complétion d'une charge récurrente..."
+      label: Proposed Solution & User Experience (UX)
+      description: Describe precisely the desired solution, user interactions, and expected outcome.
+      placeholder: "Example: Introduce an automatic alternating rotation rule upon completion of each recurring task..."
     validations:
       required: true
   - type: textarea
     id: impact-analytics
     attributes:
-      label: Mesure d'impact & Analytics souhaités
-      description: Quels événements télémétriques (event_name, paramètres) ou métriques d'usage permettront de mesurer l'adoption et le succès de cette fonctionnalité ?
-      placeholder: "Exemple :\n- event_name: 'recurring_task_created' (params: frequency, is_duo_rotating, area)\n- event_name: 'duo_rotation_assigned' (params: frequency, rotation_role, cycle_count)"
+      label: Desired Impact Metrics & Telemetry
+      description: Which telemetry events (event_name, parameters) or usage metrics will measure adoption and success?
+      placeholder: "Example:\n- event_name: 'recurring_task_created' (params: frequency, is_duo_rotating, area)\n- event_name: 'duo_rotation_assigned' (params: frequency, rotation_role, cycle_count)"
     validations:
       required: false
   - type: textarea
     id: technical-architecture
     attributes:
-      label: Pistes Techniques & Architecture (Optionnel)
-      description: Éléments pressentis pour l'implémentation (Room Schema, StateFlow, ViewModel, Jetpack Compose, Gemini, Firestore).
-      placeholder: "Exemple : Ajout d'un champ `rotationRule` dans MentalLoadItem, mise à jour de SecondBrainRepository..."
+      label: Technical Architecture & Implementation Notes (Optional)
+      description: Anticipated implementation details (Room Schema, StateFlow, ViewModel, Jetpack Compose, Gemini, Firestore).
+      placeholder: "Example: Add 'rotationRule' field to MentalLoadItem, update SecondBrainRepository..."
     validations:
       required: false
   - type: checkboxes
     id: quality-commitments
     attributes:
-      label: Engagements Qualité & Guidelines
-      description: Critères obligatoires de validation pour l'implémentation future.
+      label: Quality Commitments & Guidelines
+      description: Mandatory validation criteria for future implementation.
       options:
-        - label: Respect de l'architecture MVI / Clean Kotlin (ViewModel, StateFlow, Room / Firestore)
+        - label: Adherence to MVI / Clean Kotlin architecture (ViewModel, StateFlow, Room / Firestore)
           required: true
-        - label: UI 100% Jetpack Compose conforme aux principes Serene UI (animations douces, zéro friction)
+        - label: 100% Jetpack Compose UI compliant with Serene UI principles (gentle animations, zero friction)
           required: true
-        - label: Contrat d'Analytics & Observabilité formalisé
+        - label: Formalized Analytics & Observability contract
           required: true
-        - label: Tests unitaires & Robolectric (100% passing sur `./scripts/quality-check.sh`)
+        - label: Unit tests & Robolectric (100% passing on './scripts/quality-check.sh')
           required: true
diff --git a/README.md b/README.md
@@ -22,24 +22,24 @@ Autonomous AI coding agents often derail when given unbounded write access to pr
 ```mermaid
 flowchart TD
     subgraph Inception["Phase 1 · Inception Consortium"]
-        P1["P1: Product Planner\nAnti-duplication & User Stories Gherkin"] --> P2["P2: Design Lead\nTokens M3, UI Matrix 4 états"]
-        P2 --> P3["P3: Privacy & Data\nZero-PII, Taxonomie analytique, RGPD"]
-        P3 --> P4["P4: System Architect\nClean MVI, Boundaries, Découpage DAG Epics"]
+        P1["P1: Product Planner\nAnti-duplication & Gherkin User Stories"] --> P2["P2: Design Lead\nM3 Tokens, 4-State UI Matrix"]
+        P2 --> P3["P3: Privacy & Data\nZero-PII, Analytics Taxonomy, GDPR"]
+        P3 --> P4["P4: System Architect\nClean MVI, Boundaries, Epic DAG Decomposition"]
         P4 --> Spec["Dual-Write Spec\nGitHub Issue Comment + implementation_plan.md"]
     end
 
-    Spec --> Gate14{"🛑 GATE 1.4 : Inception Halt\nApprobation explicite écrite\n(Branch Guard actif sur main/master)"}
+    Spec --> Gate14{"🛑 GATE 1.4 : Inception Halt\nExplicit Written Approval\n(Branch Guard active on default branch)"}
 
     subgraph Delivery["Phase 2 · Delivery"]
-        Gate14 -->|Validation explicite| Branch["Création de branche dédiée\n<type>/issue-<id>-<slug>"]
-        Branch --> P5["P5: Software Engineer\nImplémentation TDD & Compose strict\nZéro chaîne en dur (strings.xml)"]
+        Gate14 -->|Explicit approval| Branch["Dedicated Feature Branch\n<type>/issue-<id>-<slug>"]
+        Branch --> P5["P5: Software Engineer\nTDD & Strict Compose Implementation\nZero Hardcoded Strings (strings.xml)"]
         P5 --> Airbag["🛡️ Quality Airbag\n./scripts/quality-check.sh\n(codeSanityCheck + Lint + Roborazzi)"]
     end
 
     subgraph Release["Phase 3 · Release"]
-        Airbag -->|Vérifications 100% vertes| P6["P6: Release Manager\nPush git & Ouverture PR Walkthrough"]
-        P6 --> Gate35{"🔒 GATE 3.5 : Auto-Merge Lock\nAttente formelle du développeur\n('Tu peux merger')"}
-        Gate35 -->|Confirmation écrite| Merge["Squash Merge & Dual-Sync\nClôture Kanban & Pruning local"]
+        Airbag -->|100% green checks| P6["P6: Release Manager\nGit Push & Open PR Walkthrough"]
+        P6 --> Gate35{"🔒 GATE 3.5 : Auto-Merge Lock\nHuman Authorization Required\n('Tu peux merger')"}
+        Gate35 -->|Written confirmation| Merge["Squash Merge & Dual-Sync\nKanban Closure & Branch Pruning"]
     end
 
     style Gate14 fill:#ff4d4f,stroke:#333,stroke-width:2px,color:#fff

diff --git a/scripts/README.md b/scripts/README.md
@@ -1,33 +1,32 @@
-# 🛠️ Scripts d'Automatisation — Agentic Android Kernel
+# 🛠️ Automation Scripts — Agentic Android Kernel
 
-Ce dossier contient l'ensemble des **scripts atomiques** et des **orchestrateurs** du projet **Agentic Android Kernel**.
+This directory contains all **atomic scripts** and **orchestrators** for the **Agentic Android Kernel** project.
 
-> 📖 **Documentation Complète & Architecture Détaillée :**
-> Consultez [**`docs/scripts-reference.md`**](../docs/scripts-reference.md) pour les fiches exhaustives, les options CLI, et la matrice de composition.
+> 📖 **Comprehensive Documentation & Detailed Architecture:**  
+> See [**`docs/scripts-reference.md`**](../docs/scripts-reference.md) for exhaustive reference sheets, CLI options, and the composition matrix.
 
 ---
 
-## ⚡ Index Rapide des Scripts
+## ⚡ Quick Script Index
 
-| Script | Langage | Description Synthétique | Commande |
+| Script | Language | Concise Description | Command |
 |---|---|---|---|
-| [`validate-docs.sh`](validate-docs.sh) | Bash | Valide la présence et la conformité des contrats de spécifications (`DESIGN.md`, playbooks, skills, règles) | `./scripts/validate-docs.sh` |
-| [`quality-check.sh`](quality-check.sh) | Bash | Exécute la suite complète de compilation Kotlin/Java, Lint Android et tests unitaires/UI | `./scripts/quality-check.sh` |
-| [`test-firestore-rules.mjs`](test-firestore-rules.mjs) | Node.js | Exécute 16 assertions de sécurité et de non-régression sur `firestore.rules` | `node scripts/test-firestore-rules.mjs` |
-| [`deploy-firestore-rules.mjs`](deploy-firestore-rules.mjs) | Node.js | Déploie et publie `firestore.rules` via l'API REST avec JWT compte de service | `node scripts/deploy-firestore-rules.mjs` |
-| [`deploy-app-distribution.sh`](deploy-app-distribution.sh) | Bash | Compile et déploie l'APK sur Firebase App Distribution avec multiplicateur de version (x100) | `./scripts/deploy-app-distribution.sh` |
-| [`generate-screenshots.sh`](generate-screenshots.sh) | Bash | Génère les captures d'écran Roborazzi pour le Design System | `./scripts/generate-screenshots.sh` |
-| [`upload-screenshots.py`](upload-screenshots.py) | Python | Valide et mappe les captures Roborazzi sur les écrans Google Stitch | `python3 scripts/upload-screenshots.py --check-only` |
-| [`cleanup-e2e-firestore.mjs`](cleanup-e2e-firestore.mjs) | Node.js | Purge les documents de sondes de test dans la collection Firestore `/users` | `node scripts/cleanup-e2e-firestore.mjs` |
-| [`inspect-ide.sh`](inspect-ide.sh) | Bash | Exécute les inspections Android Studio / IntelliJ en mode headless | `./scripts/inspect-ide.sh` |
+| [`validate-docs.sh`](validate-docs.sh) | Bash | Validates presence and contract conformance of specifications (`DESIGN.md`, playbooks, skills, rules) | `./scripts/validate-docs.sh` |
+| [`quality-check.sh`](quality-check.sh) | Bash | Executes the complete Kotlin/Java compilation, Android Lint, and unit/UI test suite | `./scripts/quality-check.sh` |
+| [`test-runtime-guardrails.mjs`](test-runtime-guardrails.mjs) | Node.js | Runs 18 security assertions on runtime guardrails (`branch-guard.mjs`, `plan-guard.mjs`) | `node scripts/test-runtime-guardrails.mjs` |
+| [`install-hooks.sh`](install-hooks.sh) | Bash | Binds local Git hooks to `.agent/hooks/` and enables the pre-commit airbag | `./scripts/install-hooks.sh` |
+| [`deploy-app-distribution.sh`](deploy-app-distribution.sh) | Bash | Builds and deploys APK to Firebase App Distribution with monotonic version multiplier (x100) | `./scripts/deploy-app-distribution.sh` |
+| [`generate-screenshots.sh`](generate-screenshots.sh) | Bash | Generates Roborazzi screenshots for the Design System | `./scripts/generate-screenshots.sh` |
+| [`upload-screenshots.py`](upload-screenshots.py) | Python | Validates and maps Roborazzi screenshots to Google Stitch screens | `python3 scripts/upload-screenshots.py --check-only` |
+| [`inspect-ide.sh`](inspect-ide.sh) | Bash | Runs Android Studio / IntelliJ inspections in headless mode | `./scripts/inspect-ide.sh` |
 
 ---
 
-## 🎯 Composition par Compétences Sémantiques ([`.agent/skills/`](../.agent/skills/))
+## 🎯 Composition by Semantic Skills ([`.agent/skills/`](../.agent/skills/))
 
-Chaque script remplit une responsabilité unique. Les enchaînements sont pilotés par les compétences sémantiques natives de l'agent :
-- **`quality-airbag`** (`/quality-check`) ➔ `validate-docs.sh` + `quality-check.sh` + `test-firestore-rules.mjs`
+Each script fulfills a single responsibility. End-to-end workflows are driven by the agent's native semantic skills:
+- **`quality-airbag`** (`/quality-check`) ➔ `validate-docs.sh` + `quality-check.sh` + `test-runtime-guardrails.mjs`
 - **`sync-stitch`** (`/sync-stitch`) ➔ `validate-docs.sh` + `generate-screenshots.sh` + `upload-screenshots.py`
 - **`distribute-local`** (`/distribute-local`) ➔ `quality-check.sh` + `deploy-app-distribution.sh`
 - **`open-pr`** (`/open-pr`) ➔ `quality-check.sh` + Walkthrough & PR creation
-- **`triage-feedback`** (`/triage-feedback`) ➔ 100% MCP natif (`GitHubMCP` : `search_issues`, `add_issue_comment`, `create_issue`)
+- **`triage-feedback`** (`/triage-feedback`) ➔ 100% native MCP (`GitHubMCP`: `search_issues`, `add_issue_comment`, `create_issue`)