erne-universal 0.5.1 → 0.6.1

package/CLAUDE.md

@@ -46,9 +46,60 @@ You are working in an ERNE-powered React Native/Expo project.
 - Branch naming: `feat/`, `fix/`, `refactor/` prefix
 - Atomic commits — one logical change per commit
 
+## Workflow Orchestration
+
+### 1. Plan Mode Default
+- Enter plan mode for ANY non-trivial task (3+ steps or architectural decisions)
+- If something goes sideways, STOP and re-plan immediately — don't keep pushing
+- Use plan mode for verification steps, not just building
+- Write detailed specs upfront to reduce ambiguity
+
+### 2. Subagent Strategy
+- Use subagents liberally to keep main context window clean
+- Offload research, exploration, and parallel analysis to subagents
+- For complex problems, throw more compute at it via subagents
+- One task per subagent for focused execution
+
+### 3. Self-Improvement Loop
+- After ANY correction from the user: save a `feedback` memory via the memory system
+- Write rules for yourself that prevent the same mistake
+- Memory is auto-loaded each session — no manual review needed
+
+### 4. Verification Before Done
+- Never mark a task complete without proving it works
+- Diff behavior between main and your changes when relevant
+- Ask yourself: "Would a staff engineer approve this?"
+- Run tests, check logs, demonstrate correctness
+
+### 5. Demand Elegance (Balanced)
+- For non-trivial changes: pause and ask "is there a more elegant way?"
+- If a fix feels hacky: "Knowing everything I know now, implement the elegant solution"
+- Skip this for simple, obvious fixes — don't over-engineer
+- Challenge your own work before presenting it
+
+### 6. Autonomous Bug Fixing
+- When given a bug report: just fix it. Don't ask for hand-holding
+- Point at logs, errors, failing tests — then resolve them
+- Zero context switching required from the user
+- Go fix failing CI tests without being told how
+
+## Task Management
+
+1. **Plan First**: Use plan mode or TodoWrite for non-trivial tasks
+2. **Verify Plan**: Check in before starting implementation
+3. **Track Progress**: Use built-in task system (TodoWrite/TaskCreate) to mark progress
+4. **Explain Changes**: High-level summary at each step
+5. **Capture Lessons**: Save corrections as `feedback` memories
+
+## Core Principles
+
+- **Simplicity First**: Make every change as simple as possible. Impact minimal code.
+- **No Laziness**: Find root causes. No temporary fixes. Senior developer standards.
+- **Minimal Impact**: Changes should only touch what's necessary. Avoid introducing bugs.
+
 ## Available Commands
 
-Use `/plan`, `/code-review`, `/tdd`, `/build-fix`, `/perf`, `/upgrade`, `/native-module`, `/navigate`, `/animate`, `/deploy`, `/component`, `/debug`, `/quality-gate` for guided workflows.
+Use `/plan`, `/code-review`, `/tdd`, `/build-fix`, `/perf`, `/upgrade`, `/native-module`, `/navigate`, `/animate`, `/deploy`, `/component`, `/debug`, `/quality-gate`, `/code`, `/feature`, `/learn`, `/retrospective`, `/setup-device` for guided workflows.
 
 ## Available Skills
 

package/README.md

@@ -1,55 +1,53 @@
-# everything-react-native-expo (ERNE)
+# 🏗️ ERNE — Everything React Native & Expo
 
-Complete AI coding agent harness for React Native and Expo development.
+> **The complete AI coding agent harness for React Native and Expo** — 11 specialized agents, adaptive stack detection, multi-agent orchestration, and a pixel-art dashboard. Every config is generated to match *your* exact project setup.
+
+[![npm version](https://img.shields.io/npm/v/erne-universal?color=cc3534&label=npm)](https://www.npmjs.com/package/erne-universal)
+[![GitHub stars](https://img.shields.io/github/stars/JubaKitiashvili/everything-react-native-expo?style=social)](https://github.com/JubaKitiashvili/everything-react-native-expo)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+[![PRs Welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg)](https://makeapullrequest.com)
 
 <p align="center">
-  <video src="demo-dashboard.mp4" autoplay loop muted playsinline width="800"></video>
+  <video src="https://github.com/JubaKitiashvili/everything-react-native-expo/raw/main/demo-dashboard.mp4" autoplay loop muted playsinline width="800"></video>
 </p>
 
-## Quick Start
+---
+
+## ⚡ Quick Start
 
 ```bash
 npx erne-universal init
 ```
 
 This will:
-1. **Deep-scan your project** — detects 15 stack dimensions (state management, navigation, styling, lists, images, forms, storage, testing, build system, component style, monorepo, New Architecture, and more)
-2. Let you choose a hook profile (minimal / standard / strict)
-3. Select MCP integrations (simulator control, GitHub, etc.)
-4. **Generate adaptive configuration** — selects from 24 variant templates matched to your exact stack (Zustand vs Redux, Expo Router vs React Navigation, NativeWind vs StyleSheet, etc.)
+1. 🔍 **Deep-scan your project** — detects 15 stack dimensions (state management, navigation, styling, lists, images, forms, storage, testing, build system, component style, monorepo, New Architecture, and more)
+2. 🎚️ Let you choose a hook profile (minimal / standard / strict)
+3. 🔌 Select MCP integrations (simulator control, GitHub, etc.)
+4. ⚙️ **Generate adaptive configuration** — selects from 24 variant templates matched to your exact stack (Zustand vs Redux, Expo Router vs React Navigation, NativeWind vs StyleSheet, etc.)
+
+---
 
-## What's Included
+## 📦 What's Included
 
 | Component | Count | Description |
 |-----------|-------|-------------|
-| Agents | 10 | Specialized AI agents for architecture, development, review, testing, UI, native, and more |
-| Agent variants | 9 | Stack-adaptive agent configurations (StyleSheet vs NativeWind, Zustand vs Redux, etc.) |
-| Commands | 16 | Slash commands for every React Native workflow |
-| Rule layers | 5 | Conditional rules: common, expo, bare-rn, native-ios, native-android |
-| Rule variants | 15 | Stack-specific rules selected by deep detection (state, navigation, styling, security, etc.) |
-| Hook profiles | 3 | Minimal, standard, strict — quality enforcement your way |
-| Skills | 8 | Reusable knowledge modules loaded on-demand |
-| Contexts | 3 | Behavior modes: dev, review, vibe |
-| MCP configs | 10 | Pre-configured server integrations |
-
-## Token Efficiency
+| 🤖 Agents | 11 | Specialized AI agents for architecture, development, review, testing, UI, native, orchestration |
+| 🔀 Agent variants | 9 | Stack-adaptive agent configurations (StyleSheet vs NativeWind, Zustand vs Redux, etc.) |
+| ⚡ Commands | 19 | Slash commands for every React Native workflow |
+| 📏 Rule layers | 5 | Conditional rules: common, expo, bare-rn, native-ios, native-android |
+| 🎯 Rule variants | 15 | Stack-specific rules selected by deep detection (state, navigation, styling, security, etc.) |
+| 🛡️ Hook profiles | 3 | Minimal, standard, strict — quality enforcement your way |
+| 📚 Skills | 7 | Reusable knowledge modules loaded on-demand |
+| 🎭 Contexts | 3 | Behavior modes: dev, review, vibe |
+| 🔌 MCP configs | 10 | Pre-configured server integrations |
+| 📋 Workflow examples | 4 | End-to-end multi-agent workflow guides |
+| 🤝 Handoff templates | 4 | Structured agent-to-agent context passing |
 
-ERNE's architecture is designed to minimize token usage through six layered mechanisms:
+---
 
-| Mechanism | How it works | Savings |
-|-----------|-------------|---------|
-| **Profile-gated hooks** | Minimal profile runs 3 hooks instead of 17 | ~31% |
-| **Conditional rules** | Only loads rules matching your project type (Expo, bare RN, native) | ~26% |
-| **On-demand skills** | Skills load only when their command is invoked, not always in context | ~12% |
-| **Subagent isolation** | Fresh agent per task with only its own definition + relevant rules | ~12% |
-| **Task-specific commands** | 16 focused prompts instead of one monolithic instruction set | ~13% |
-| **Context-based behavior** | Modes change behavior dynamically without loading new rulesets | ~3% |
-
-**Result:** Typical workflows use **60–67% fewer tokens** compared to a naive all-in-context approach. Vibe mode (minimal profile) reaches 67% savings, standard development 64%, and even strict mode saves 57%.
+## 🎮 Agent Dashboard
 
-## Agent Dashboard
-
-ERNE includes a real-time pixel-art dashboard that visualizes all 10 agents working in an animated office environment.
+ERNE includes a real-time pixel-art dashboard that visualizes all 11 agents working in an animated office environment.
 
 ```bash
 erne dashboard              # Start on port 3333, open browser
@@ -59,69 +57,122 @@ erne start                  # Init project + dashboard in background
 ```
 
 **Features:**
-- 4 office rooms — Development, Code Review, Testing, and Conference (brainstorming)
-- 10 animated agent sprites with walking, typing, and done animations
-- Conference room brainstorming — agents gather for planning sessions
-- Thought bubbles showing the current task above working agents
-- Animated monitor screens (green code when working, screensaver when idle)
-- Toast notifications for agent start/complete events
-- Bottom stats bar with session duration, tasks completed, working/planning counts
-- Agent detail overlay with full activity history (click any agent)
-- Office decorations: plants, coffee machine, water cooler, bookshelves, posters, wall clocks
-- Real-time status updates via WebSocket (connected to Claude Code hooks)
-- Auto-reconnect with exponential backoff
-
-The dashboard hooks into Claude Code's `PreToolUse` and `PostToolUse` events (pattern: `Agent`) to track which agents are actively working and what they're doing.
-
-## IDE & Editor Support
+- 🏢 4 office rooms — Development, Code Review, Testing, and Conference
+- 🎨 11 animated agent sprites with walking, typing, and done animations
+- 💬 Thought bubbles showing the current task above working agents
+- 🖥️ Animated monitor screens (green code when working, screensaver when idle)
+- 🔔 Toast notifications for agent start/complete events
+- 📊 Bottom stats bar with session duration, tasks completed, working/planning counts
+- 👤 Agent detail overlay with full activity history (click any agent)
+- 🎯 Pipeline orchestrator coordination view in conference room
+- 🔄 Real-time WebSocket updates with auto-reconnect
 
-ERNE works with every major AI coding assistant out of the box:
+---
 
-| File | IDE / Tool |
-|------|-----------|
-| `CLAUDE.md` | Claude Code |
-| `AGENTS.md` | Codex, Windsurf, Cursor, GitHub Copilot |
-| `GEMINI.md` | Google Antigravity |
-| `.cursorrules` | Cursor |
-| `.windsurfrules` | Windsurf |
-| `.github/copilot-instructions.md` | GitHub Copilot |
+## 🎯 Multi-Agent Orchestration
+
+ERNE supports coordinated multi-agent workflows through the pipeline orchestrator:
+
+```bash
+/orchestrate "build user profile screen"
+```
+
+**Pipeline phases:**
+
+```
+  🏗️ Plan          architect decomposes the task
+       ↓
+  ⚡ Implement     senior-developer + feature-builder (parallel)
+       ↓
+  🧪 Test          tdd-guide writes and runs tests
+       ↓
+  🔍 Review        code-reviewer validates with evidence
+       ↓
+  📊 Validate      performance-profiler checks performance
+```
+
+Features retry logic (max 3 attempts), escalation to user on persistent failures, and structured [handoff templates](docs/handoff-templates.md) for context passing between agents. See [Pipeline Documentation](docs/pipeline.md) for details.
+
+---
+
+## 🤖 Agents
+
+Each agent has a distinct personality, quantified success metrics, and memory integration for cross-session learning.
+
+| Agent | Emoji | Domain | Room |
+|-------|-------|--------|------|
+| **architect** | 🏗️ | System design and project structure | Development |
+| **senior-developer** | 👨‍💻 | End-to-end feature implementation, screens, hooks, API | Development |
+| **feature-builder** | ⚡ | Focused implementation units, works in parallel | Development |
+| **native-bridge-builder** | 🌉 | Turbo Modules and native platform APIs | Development |
+| **expo-config-resolver** | ⚙️ | Expo configuration and build issues | Development |
+| **ui-designer** | 🎨 | Accessible, performant UI components | Development |
+| **code-reviewer** | 🔍 | Code quality with evidence-based approval | Code Review |
+| **upgrade-assistant** | 📦 | Version migration guidance | Code Review |
+| **tdd-guide** | 🚦 | Test-driven development workflow | Testing |
+| **performance-profiler** | 🏎️ | FPS diagnostics and bundle optimization | Testing |
+| **pipeline-orchestrator** | 🎯 | Multi-agent workflow coordination | Conference |
 
-All config files are generated adaptively based on your project's detected stack — state management, navigation, styling, and more — so the AI guidance matches your actual setup.
+---
 
-## Agents
+## 💰 Token Efficiency
 
-| Agent | Domain | Room |
-|-------|--------|------|
-| **architect** | System design and project structure | Development |
-| **senior-developer** | End-to-end feature implementation, screens, hooks, API | Development |
-| **feature-builder** | Focused implementation units, works in parallel | Development |
-| **native-bridge-builder** | Turbo Modules and native platform APIs | Development |
-| **expo-config-resolver** | Expo configuration and build issues | Development |
-| **ui-designer** | Accessible, performant UI components | Development |
-| **code-reviewer** | Code quality and best practices | Code Review |
-| **upgrade-assistant** | Version migration guidance | Code Review |
-| **tdd-guide** | Test-driven development workflow | Testing |
-| **performance-profiler** | FPS diagnostics and bundle optimization | Testing |
+ERNE's architecture is designed to minimize token usage through six layered mechanisms:
+
+| Mechanism | How it works | Savings |
+|-----------|-------------|---------|
+| **Profile-gated hooks** | Minimal profile runs 4 hooks instead of 16 | ~31% |
+| **Conditional rules** | Only loads rules matching your project type (Expo, bare RN, native) | ~26% |
+| **On-demand skills** | Skills load only when their command is invoked, not always in context | ~12% |
+| **Subagent isolation** | Fresh agent per task with only its own definition + relevant rules | ~12% |
+| **Task-specific commands** | 19 focused prompts instead of one monolithic instruction set | ~13% |
+| **Context-based behavior** | Modes change behavior dynamically without loading new rulesets | ~3% |
+
+**Result:** Typical workflows use **60–67% fewer tokens** compared to a naive all-in-context approach.
 
-## Hook Profiles
+---
+
+## 🛡️ Hook Profiles
 
 | Profile | Hooks | Use Case |
 |---------|-------|----------|
-| minimal | 3 | Fast iteration, vibe coding — maximum speed, minimum friction |
-| standard | 11 | Balanced quality + speed (recommended) — catches real issues |
-| strict | 17 | Production-grade enforcement — full security, accessibility, perf budgets |
+| `minimal` | 4 | ⚡ Fast iteration, vibe coding — maximum speed, minimum friction |
+| `standard` | 12 | ⚖️ Balanced quality + speed (recommended) — catches real issues |
+| `strict` | 16 | 🔒 Production-grade enforcement — full security, accessibility, perf budgets |
 
 Change profile: set `ERNE_PROFILE` env var, add `<!-- Hook Profile: standard -->` to CLAUDE.md, or use `/vibe` context.
 
-## Commands
+---
+
+## ⚡ Commands
+
+| Category | Commands |
+|----------|----------|
+| **Core** | `/plan`, `/code-review`, `/tdd`, `/build-fix`, `/perf`, `/upgrade`, `/native-module`, `/navigate`, `/code`, `/feature` |
+| **Extended** | `/animate`, `/deploy`, `/component`, `/debug`, `/quality-gate` |
+| **Orchestration** | `/orchestrate` |
+| **Learning** | `/learn`, `/retrospective`, `/setup-device` |
+
+---
 
-**Core:** `/plan`, `/code-review`, `/tdd`, `/build-fix`, `/perf`, `/upgrade`, `/native-module`, `/navigate`
+## 🖥️ IDE & Editor Support
+
+ERNE works with every major AI coding assistant out of the box:
+
+| File | IDE / Tool |
+|------|-----------|
+| `CLAUDE.md` | Claude Code |
+| `AGENTS.md` | Codex, Windsurf, Cursor, GitHub Copilot |
+| `GEMINI.md` | Google Antigravity |
+| `.cursorrules` | Cursor |
+| `.windsurfrules` | Windsurf |
+| `.github/copilot-instructions.md` | GitHub Copilot |
 
-**Extended:** `/animate`, `/deploy`, `/component`, `/debug`, `/quality-gate`
+All config files are generated adaptively based on your project's detected stack.
 
-**Learning:** `/learn`, `/retrospective`, `/setup-device`
+---
 
-## Architecture
+## 🏗️ Architecture
 
 ```
 Claude Code Hooks ──▶ run-with-flags.js ──▶ Profile gate ──▶ Hook scripts
@@ -139,30 +190,32 @@ Claude Code PostToolUse ┘
 ```
 
 **Key design principles:**
-- **Zero runtime dependencies** for the harness itself (ws package only for dashboard)
-- **Conditional loading** — rules, skills, and hooks load based on project type and profile
-- **Fresh subagent per task** — no context pollution between agent invocations
-- **Silent failure** — hooks never block Claude Code if something goes wrong
+- 🪶 **Zero runtime dependencies** for the harness itself (ws package only for dashboard)
+- 🎯 **Conditional loading** — rules, skills, and hooks load based on project type and profile
+- 🧹 **Fresh subagent per task** — no context pollution between agent invocations
+- 🔇 **Silent failure** — hooks never block Claude Code if something goes wrong
 
-## Contributing
+---
+
+## 🤝 Contributing
 
 We welcome contributions from everyone — from typo fixes to new agents and skills.
 
 | I want to... | Start here |
 |--------------|-----------|
-| Report a bug | [Bug Report](https://github.com/JubaKitiashvili/everything-react-native-expo/issues/new?template=bug_report.md) |
-| Request a feature | [Feature Request](https://github.com/JubaKitiashvili/everything-react-native-expo/issues/new?template=feature_request.md) |
-| Propose a new skill | [Skill Proposal](https://github.com/JubaKitiashvili/everything-react-native-expo/issues/new?template=new_skill.md) |
-| Submit a PR | [Contributing Guide](.github/CONTRIBUTING.md) |
-
-**Quick contribution ideas:** Add a variant template for your stack, improve an agent's knowledge, write a new slash command, add an MCP config for your favorite service.
+| 🐛 Report a bug | [Bug Report](https://github.com/JubaKitiashvili/everything-react-native-expo/issues/new?template=bug_report.md) |
+| 💡 Request a feature | [Feature Request](https://github.com/JubaKitiashvili/everything-react-native-expo/issues/new?template=feature_request.md) |
+| 📚 Propose a new skill | [Skill Proposal](https://github.com/JubaKitiashvili/everything-react-native-expo/issues/new?template=new_skill.md) |
+| 🔀 Submit a PR | [Contributing Guide](CONTRIBUTING.md) |
 
 ```bash
 git checkout -b feat/your-feature
 npm run validate && npm test   # Must pass before PR
 ```
 
-## Partnerships
+---
+
+## 🤝 Partnerships
 
 Skills, agents, and MCP configs are open source — anyone can add them via PR. Partnerships are for deeper collaboration:
 
@@ -175,26 +228,45 @@ Skills, agents, and MCP configs are open source — anyone can add them via PR.
 
 If you maintain a React Native library, Expo tool, or developer service — [let's talk](https://github.com/JubaKitiashvili/everything-react-native-expo/issues/new?template=partnership.md).
 
-## Available On
+---
+
+## 📦 Available On
 
 - [npm](https://www.npmjs.com/package/erne-universal) — `npx erne-universal init`
 - [SkillsMP](https://skillsmp.com) — Auto-indexed from GitHub
 - [BuildWithClaude](https://buildwithclaude.com) — Plugin directory
 - [VoltAgent/awesome-agent-skills](https://github.com/VoltAgent/awesome-agent-skills) — Curated skills list
 
-## Documentation
+---
+
+## 📖 Documentation
+
+| Doc | Description |
+|-----|-------------|
+| [Getting Started](docs/getting-started.md) | Installation and first run |
+| [Agents Guide](docs/agents.md) | All 11 agents with domains and usage |
+| [Commands Reference](docs/commands.md) | All 19 slash commands |
+| [Hooks & Profiles](docs/hooks-profiles.md) | Hook system and 3 profiles |
+| [Creating Skills](docs/creating-skills.md) | Author your own skills |
+| [Pipeline & Orchestration](docs/pipeline.md) | Multi-agent workflow coordination |
+| [Memory Integration](docs/memory-integration.md) | Cross-session learning with MCP |
+| [Handoff Templates](docs/handoff-templates.md) | Structured agent-to-agent context |
+| [Contributing](CONTRIBUTING.md) | How to contribute |
+
+---
+
+## 📜 License
+
+MIT License — use freely, commercially or personally.
+
+---
 
-- [Getting Started](docs/getting-started.md)
-- [Agents Guide](docs/agents.md)
-- [Commands Reference](docs/commands.md)
-- [Hooks & Profiles](docs/hooks-profiles.md)
-- [Creating Skills](docs/creating-skills.md)
+<div align="center">
 
-## Links
+**🏗️ ERNE — Your React Native AI Dream Team 🏗️**
 
-- Website: [erne.dev](https://erne.dev)
-- npm: [erne-universal](https://www.npmjs.com/package/erne-universal)
+[⭐ Star this repo](https://github.com/JubaKitiashvili/everything-react-native-expo) · [🍴 Fork it](https://github.com/JubaKitiashvili/everything-react-native-expo/fork) · [🐛 Report an issue](https://github.com/JubaKitiashvili/everything-react-native-expo/issues) · [🌐 erne.dev](https://erne.dev)
 
-## License
+Made with ❤️ for the React Native community
 
-MIT
+</div>

package/agents/architect.md

@@ -1,5 +1,7 @@
 ---
 name: architect
+emoji: "\U0001F3D7\uFE0F"
+vibe: "Sees the whole board before making a move"
 description: Feature decomposition, navigation design, state management selection, API layer planning. Triggered by /plan and /navigate.
 ---
 
@@ -9,6 +11,30 @@ You are the ERNE Architect agent — a senior React Native/Expo systems designer
 
 Design feature architectures, navigation flows, and system structure for React Native and Expo applications.
 
+## Identity & Personality
+
+Methodical and strategic. You see systems as living organisms — every component affects every other. You are the one who asks "what happens when this grows 10x?" before a single line of code is written. You would rather spend an hour on a whiteboard than a week refactoring a wrong abstraction. Silence from you means you are thinking, not agreeing.
+
+## Communication Style
+
+- Lead with the constraint, not the solution — "We have 3 screens sharing user state, so we need..." not "Use Zustand"
+- Use diagrams and data flow descriptions — show how data moves before showing where files live
+- Name trade-offs explicitly — never hide complexity behind "it depends"
+
+## Success Metrics
+
+- Component count per screen: <15
+- Prop drilling depth: <3 levels
+- Bundle impact per feature: <50KB
+- Zero circular dependencies in the designed architecture
+- Every architecture plan includes a clear data flow diagram
+
+## Learning & Memory
+
+- Remember navigation structures that scaled well vs. ones that required refactoring
+- Track which state management patterns led to prop drilling complaints
+- Note which feature decompositions were too granular vs. too monolithic
+
 ## Capabilities
 
 - **Feature decomposition**: Break complex features into implementable units with clear interfaces
@@ -38,6 +64,40 @@ Design feature architectures, navigation flows, and system structure for React N
 - Consider platform differences upfront (iOS vs Android UX conventions)
 - Account for the hook profile — suggest which hooks will run on the new code
 
+## Memory Integration
+
+### What to Save
+- Architecture decisions: state management choices, navigation structure, dependency selections
+- Feature decompositions that scaled well vs. ones that required rework
+- Data flow patterns that caused issues downstream (circular deps, prop drilling)
+- Navigation structures and deep link schemas designed for the project
+
+### What to Search
+- Past architecture decisions for the same project before designing new features
+- Previous navigation structures to maintain consistency
+- Performance profiler findings that should influence architecture choices
+- Code reviewer findings about recurring structural issues
+
+### Tag Format
+```
+[architect, {project}, architecture-decisions]
+[architect, {project}, navigation-plan]
+```
+
+### Examples
+**Save** after designing a feature:
+```
+save_observation(
+  content: "Payments feature: Zustand store for cart state, TanStack Query for product catalog. Stack navigator nested in tab 3. Deep link: myapp://checkout?cartId=X",
+  tags: ["architect", "my-app", "architecture-decisions"]
+)
+```
+
+**Search** before starting a new plan:
+```
+search(query: "navigation structure tab layout", tags: ["architect", "my-app"])
+```
+
 ## Output Format
 
 ```markdown

package/agents/code-reviewer.md

@@ -1,5 +1,7 @@
 ---
 name: code-reviewer
+emoji: "\U0001F50D"
+vibe: "Trust but verify — with evidence"
 description: Re-render detection, RN anti-pattern detection, platform parity, Expo SDK validation, accessibility audit. Triggered by /code-review, /quality-gate, /deploy.
 ---
 
@@ -9,6 +11,43 @@ You are the ERNE Code Reviewer agent — a meticulous React Native code quality
 
 Perform thorough code reviews focused on React Native-specific issues, performance pitfalls, and cross-platform correctness.
 
+## Identity & Personality
+
+Skeptical by nature, constructive by choice. You have seen too many "LGTM" reviews on code that crashed in production. Evidence first, approval second. You do not nitpick formatting — tools handle that — but you will block a merge over a missing error boundary or an unvalidated deep link parameter. Your reviews teach, not just gatekeep.
+
+## Communication Style
+
+- Always cite the specific file and line — never say "somewhere in the codebase"
+- Separate blocking issues from suggestions — "Must fix" vs. "Consider"
+- Explain the failure scenario, not just the rule — "This will crash when the API returns null because..."
+
+## Success Metrics
+
+- 0 P0 bugs that make it past review into production
+- False positive rate <10% — issues flagged are real issues
+- Every review includes at least one "Positive" callout for good patterns
+- Review turnaround within the same session
+
+## Learning & Memory
+
+- Remember recurring issues per developer or per module — adapt review focus accordingly
+- Track which review comments led to actual bug prevention vs. unnecessary churn
+- Note which anti-patterns keep reappearing despite previous reviews
+
+## Evidence Requirements
+
+```
+Before approving any change:
+- [ ] All existing tests pass (verified, not assumed)
+- [ ] New code has corresponding test coverage
+- [ ] UI changes include iOS + Android screenshots
+- [ ] Performance impact measured (not estimated)
+- [ ] No console.log or debug code remaining
+- [ ] Accessibility audit passed (if UI change)
+
+Default stance: NEEDS IMPROVEMENT until proven otherwise.
+```
+
 ## Review Checklist
 
 ### 1. Re-render Detection
@@ -51,6 +90,40 @@ Perform thorough code reviews focused on React Native-specific issues, performan
 - WebView `originWhitelist` configured
 - Input sanitization on user-facing forms
 
+## Memory Integration
+
+### What to Save
+- Recurring anti-patterns found across multiple reviews (with module/developer context)
+- Review comments that prevented actual production bugs
+- False positives to avoid flagging the same non-issue in future reviews
+- Module-specific known issues and accepted trade-offs
+
+### What to Search
+- Past review findings for the module under review
+- Performance profiler baselines to catch regressions during review
+- Architecture decisions to verify implementations match the plan
+- Upgrade history to check for deprecated API usage
+
+### Tag Format
+```
+[code-reviewer, {project}, review-findings]
+[code-reviewer, {project}, test-plan]
+```
+
+### Examples
+**Save** after finding a recurring issue:
+```
+save_observation(
+  content: "features/notifications module: 3 PRs in a row had unvalidated deep link params in NotificationHandler.tsx. Need input validation wrapper or lint rule.",
+  tags: ["code-reviewer", "my-app", "review-findings"]
+)
+```
+
+**Search** before reviewing a module:
+```
+search(query: "review findings notifications module", tags: ["code-reviewer", "my-app"])
+```
+
 ## Output Format
 
 Group findings by severity: