🏗️ ERNE — Everything React Native & Expo

AI agent harness with 13 specialized agents, autonomous worker, video debugging, adaptive fix, and 29 auto-updating knowledge rules. Adaptive stack detection, multi-agent orchestration, and a pixel-art dashboard. Every config is generated to match your exact project setup.

⚡ Quick Start

Prerequisite: Claude Code is required for the full experience (13 agents, pipeline orchestration, dashboard, hooks). Other IDEs receive adaptive rules and configuration — see IDE Support.

Install & Setup

# One-time setup in your React Native / Expo project
npx erne-universal init

# Launch the dashboard
npx erne-universal dashboard

# Or do both at once
npx erne-universal start

What init does:

1. 🔍 Deep-scans 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. 🎚️ Lets you choose a hook profile (minimal / standard / strict)
3. 🔌 Select MCP integrations (simulator control, GitHub, etc.)
4. ⚙️ Generates adaptive configuration — selects from 24 variant templates matched to your exact stack (Zustand vs Redux, Expo Router vs React Navigation, NativeWind vs StyleSheet, etc.)

CLI Commands

npx erne-universal init                  # Setup ERNE in your project
npx erne-universal dashboard             # Launch dashboard
npx erne-universal start                 # Init + dashboard
npx erne-universal doctor                # Health check (22 checks)
npx erne-universal audit                 # Generate project documentation
npx erne-universal worker --config w.json  # Start autonomous worker
npx erne-universal sync-configs          # Export rules to other IDEs
npx erne-universal uninstall             # Clean removal

📦 What's Included

| Component | Count | Description | | -------------------- | ----- | -------------------------------------------------------------------------------------------- | | 🤖 Agents | 13 | Specialized AI agents incl. visual debugger, doc generator, smart routing | | 🔀 Agent variants | 9 | Stack-adaptive agent configurations (StyleSheet vs NativeWind, Zustand vs Redux, etc.) | | ⚡ Commands | 23 | Slash commands for every React Native workflow (incl. /erne-debug-video) | | 📏 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.) | | 📚 Knowledge rules | 29 | Expo SDK 55, RN 0.84, React 19.2, Reanimated, Skia, Gesture Handler, SVG, ExecuTorch, more | | 🛡️ 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 | | 🎬 Video debugging | 1 | Analyze screen recordings for temporal UI bugs | | 🔧 Adaptive Fix | — | Agent-based or direct fix from dashboard | | 📋 Workflow examples | 4 | End-to-end multi-agent workflow guides | | 🤝 Handoff templates | 4 | Structured agent-to-agent context passing |

🎮 Agent Dashboard

ERNE includes a real-time dashboard with 6 pages, pixel-art agent HQ, and adaptive fix integration.

npx erne-universal dashboard              # Start on port 3333, open browser
npx erne-universal dashboard --port 4444  # Custom port
npx erne-universal start                  # Init + dashboard in one command

Command Center

Health score, issues summary, agent status, system health checks.

Issues — Agent Fix

Severity filtering, Agent Fix buttons (auto-detects Claude Code), "Needs planning" for complex issues, real-time fix output.

Tasks — Worker & Providers

6 provider integrations (ClickUp, GitHub, Linear, Jira, GitLab, Local), Kanban board, interactive setup guides.

Agents — Pixel Art HQ

13 animated agents in 4 rooms, real-time status, custom naming, activity history.

Project — Stack & Docs

Stack detection, MCP servers, generated documentation (12 doc types), recommendations.

Commands

23 commands including /erne-debug-video, click to copy.

🎬 Video Debugging

Analyze screen recordings to find temporal UI bugs that screenshots cannot capture.

/erne-debug-video recording.mp4

Catches: animation glitches, race conditions, gesture issues, scroll jank, keyboard overlap, navigation transitions.

• Extracts key frames via ffmpeg scene detection (any format: mp4, mov, webm, avi, mkv, gif)
• Claude analyzes frames as a timeline with frame references
• No additional API keys needed — uses Claude you already have

📚 29 Knowledge Rules (Auto-Updating)

ERNE ships with 29 comprehensive rule files covering the entire React Native ecosystem:

| Category | Rules | Coverage | | ---------------- | ----- | ---------------------------------------------------------------------------------- | | Core | 10 | Expo SDK 55, RN 0.84, React 19.2, navigation, patterns, styling, testing, security | | Expo Packages | 4 | 40+ packages: audio, video, camera, file-system, sqlite, notifications, location | | Software Mansion | 5 | Reanimated v4, Gesture Handler, Skia, SVG, Screens | | Callstack | 5 | Native bottom tabs, RNTL, on-device AI, Reassure, Voltra | | Cutting-edge | 5 | ExecuTorch (on-device LLM), Audio API, Enriched (rich text), Freeze, Screens |

Rules auto-update weekly via GitHub Action — checks npm for new SDK/RN versions, analyzes changelogs with Claude API, and opens a PR.

🎯 Multi-Agent Orchestration

ERNE supports coordinated multi-agent workflows through the pipeline orchestrator:

/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 for context passing between agents. See Pipeline Documentation 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 | | visual-debugger | 🔬 | Screenshot-based UI debugging | Development | | documentation-generator | 📝 | Auto-generate project documentation | Development |

🧠 Context Optimization

ERNE includes a built-in context intelligence system (auto-enabled with dashboard) that compresses tool outputs by 97-100%, indexes content with FTS5 search, and manages your context budget. See BENCHMARK.md for the full 21-scenario breakdown.

💰 Token Efficiency

ERNE minimizes token usage through two complementary systems: architecture-level savings (what gets loaded into context) and runtime context optimization (how tool outputs and session state are compressed).

Architecture savings

| 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 | 23 focused prompts instead of one monolithic instruction set | ~13% | | Context-based behavior | Modes change behavior dynamically without loading new rulesets | ~3% |

Runtime context optimization (benchmark-verified)

| Mechanism | How it works | Savings | | ---------------------- | ------------------------------------------------------------------------ | ---------------------- | | Content summarizer | Auto-detects 14 content types, produces statistical summaries | 97–100% per output | | Index + Search | FTS5 BM25 retrieval returns only relevant chunks, code preserved exactly | 80% per search | | Smart truncation | 4-tier fallback: Structured → Pattern → Head/Tail → Hash | 85–100% per output | | Session snapshots | Captures full session state in <2KB | ~50% vs log replay | | Budget enforcement | Throttling at 80% prevents runaway token usage | Prevents overflow |

Result: Architecture saves 60–67% on what enters context. Runtime optimization achieves 97–100% compression on tool outputs (verified across 21 benchmark scenarios with 537 KB of real data). In a full debugging session, 99% of tool output tokens are eliminated — leaving 99.6% of your context window free for actual problem solving. See BENCHMARK.md for complete results.

🛡️ Hook Profiles

| Profile | Hooks | Use Case | | ---------- | ----- | ---------------------------------------------------------------------------- | | 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

| Category | Commands | | ----------------- | ---------------------------------------------------------------------------------------------------------------------- | | Core | /plan, /code-review, /tdd, /build-fix, /perf, /upgrade, /native-module, /navigate, /code, /feature | | Extended | /animate, /deploy, /component, /debug, /debug-visual, /debug-video, /quality-gate | | Orchestration | /orchestrate, /worker | | Learning | /learn, /retrospective, /setup-device |

🖥️ IDE & Editor Support

ERNE generates adaptive config files for multiple IDEs, but the full agent experience requires Claude Code:

| Feature | Claude Code | Cursor / Windsurf / Copilot / Codex | | ------------------------------- | :---------: | :---------------------------------: | | Adaptive rules & config | ✅ | ✅ | | Stack detection (15 dimensions) | ✅ | ✅ | | 23 slash commands | ✅ | ❌ | | 13 specialized agents | ✅ | ❌ | | Pipeline orchestration | ✅ | ❌ | | Hook profiles | ✅ | ❌ | | Agent dashboard | ✅ | ❌ | | Cross-session memory | ✅ | ❌ |

Generated config files:

| File | IDE / Tool | | --------------------------------- | --------------------------------------- | | CLAUDE.md | Claude Code (full experience) | | AGENTS.md | Codex, Windsurf, Cursor, GitHub Copilot | | GEMINI.md | Gemini CLI | | .cursorrules | Cursor | | .windsurfrules | Windsurf | | .github/copilot-instructions.md | GitHub Copilot |

🏗️ Architecture

Claude Code Hooks ──▶ run-with-flags.js ──▶ Profile gate ──▶ Hook scripts
                                                │
                                     ┌──────────┴──────────┐
                                     │   Only hooks for    │
                                     │   active profile    │
                                     │   are executed      │
                                     └─────────────────────┘

erne dashboard ──▶ HTTP + WS Server ──▶ Browser Canvas
                        ▲
Claude Code PreToolUse ─┤  (Agent pattern)
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

🤝 Contributing

We welcome contributions from everyone — from typo fixes to new agents and skills.

| I want to... | Start here | | ---------------------- | ------------------------------------------------------------------------------------------------------------------------- | | 🐛 Report a bug | Bug Report | | 💡 Request a feature | Feature Request | | 📚 Propose a new skill | Skill Proposal | | 🔀 Submit a PR | Contributing Guide |

git checkout -b feat/your-feature
npm run validate && npm test   # Must pass before PR

🤝 Partnerships

Skills, agents, and MCP configs are open source — anyone can add them via PR. Partnerships are for deeper collaboration:

| Partnership Type | What It Means | | -------------------- | ---------------------------------------------------------------------------------------------------- | | Co-Maintenance | You keep your library's ERNE skill up to date as your API evolves | | Early Access | We update ERNE before your breaking changes ship, so users never hit stale guidance | | Joint Promotion | Your docs recommend ERNE for AI-assisted development, we feature you on erne.dev | | Domain Expertise | Co-develop specialized agents that require deep knowledge of your platform |

If you maintain a React Native library, Expo tool, or developer service — let's talk.

📦 Available On

• npm — npx erne-universal init
• SkillsMP — Auto-indexed from GitHub
• BuildWithClaude — Plugin directory
• VoltAgent/awesome-agent-skills — Curated skills list

📖 Documentation

| Doc | Description | | ------------------------------------------------ | ------------------------------------ | | Getting Started | Installation and first run | | Agents Guide | All 13 agents with domains and usage | | Commands Reference | All 23 slash commands | | Hooks & Profiles | Hook system and 3 profiles | | Creating Skills | Author your own skills | | Pipeline & Orchestration | Multi-agent workflow coordination | | Memory Integration | Cross-session learning with MCP | | Handoff Templates | Structured agent-to-agent context | | Contributing | How to contribute |

📜 License

MIT License — use freely, commercially or personally.

🏗️ ERNE — Your React Native AI Dream Team 🏗️

⭐ Star this repo · 🍴 Fork it · 🐛 Report an issue · 🌐 erne.dev

Made with ❤️ for the React Native community