WeClaw is an open-source MCP & Skills Marketplace — discover, publish, and debug MCP servers and AI skills in real time. Connect them to your AI agents via a hosted proxy layer, all from a unified web UI.
- MCP & Skills Marketplace — Browse and search public MCP servers and AI skills by category
- Real-time Debugging — Test and debug MCP tools and skills directly in the browser with live feedback
- One-click Import — Add any
stdio/SSE/Streamable-HTTPserver from a config snippet or npm/pip package - Auto Compilation — Automatically connects to the server, introspects tools / prompts / resources, and persists the metadata
- Hosted Proxy — Wraps local
stdioservers into a remotely-accessible HTTP endpoint (no client-side runtime required) - Agent Chat — Built-in chat interface backed by CopilotKit + AG-UI; interact with your servers directly in the browser
- GitHub OAuth — Sign in with GitHub; servers can be public or private
- Admin Panel — Review, approve, and recommend servers; manage users and categories
- i18n — Chinese / English interface
| Layer | Technology |
|---|---|
| Frontend | Next.js 15, React, Tailwind CSS, Radix UI, CopilotKit |
| Backend | Python 3.10+, FastAPI, Celery, SQLAlchemy 2, Alembic |
| Database | MySQL 8 (PostgreSQL also supported) |
| Cache / Queue | Redis |
| Proxy | Caddy 2 |
| MCP Runtime | fastmcp, mcp SDK |
| Requirement | Minimum |
|---|---|
| CPU | 4 virtual cores |
| RAM | 8 GB |
| Docker | 19.03.9+ |
| Docker Compose | 1.25.1+ |
git clone https://github.com/tomeai/weclaw.git
cd weclaw
# Copy and edit the environment file
cp backend/deploy/.env.server.example backend/deploy/.env.server
# Start all services
cd docker
docker compose up -d