Menu
Add a Model Context Protocol (MCP) server to a genai-tk project — config, credential management, and validation.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
Build or modify LangChain, DeepAgent, DeerFlow profiles, agent tools, middleware, checkpointing, skills wiring, and the shared harness layer in genai-tk.
Work on BAML structured extraction, BAML CLI commands, processors, utilities, and Prefect BAML workflow integration in genai-tk.
Work on browser automation, sandbox browser tools, direct Playwright tools, AioSandbox backend, and sandbox CLI support in genai-tk.
Add or modify genai-tk Typer CLI commands, dynamic command registration, project scaffolding, and generated Copilot/agent support files.
Work on genai-tk OmegaConf configuration, profiles, overrides, env substitution, and config discovery. Use when editing config/*.yaml or genai_tk.config_mgmt.config_mngr.
Work on core LLM, embeddings, vector store, provider, cache, prompt, and retriever factories in genai-tk. Use when editing genai_tk/core or provider configuration.
| name | add-mcp-server |
| description | Add a Model Context Protocol (MCP) server to a genai-tk project — config, credential management, and validation. |
| tags | ["mcp","tools","configuration"] |
| version | 1.0 |
MCP (Model Context Protocol) servers expose tools to agents over a standardized JSON-RPC interface.
genai-tk manages them through config/mcp_servers.yaml.
Use an existing server:
@modelcontextprotocol/server-filesystem — file system access@modelcontextprotocol/server-brave-search — web search@modelcontextprotocol/server-github — GitHub APIImplement your own:
See examples/mcp_server/ for a Python MCP server skeleton.
Edit config/mcp_servers.yaml:
mcp_servers:
my_server: # ← key used in agent profiles
transport: stdio # stdio | sse
command: npx
args:
- "-y"
- "@modelcontextprotocol/server-brave-search"
env:
BRAVE_API_KEY: ${oc.env:BRAVE_API_KEY} # env var interpolation
enabled: true
my_python_server:
transport: stdio
command: uv
args:
- run
- python
- examples/mcp_server/server.py
enabled: true
In config/agents/langchain.yaml:
langchain_agents:
my_agent:
mcp_servers:
- my_server
- my_python_server
Add to .env (never commit this file):
BRAVE_API_KEY=bsk-xxxxx
Add to .env.example (commit this):
BRAVE_API_KEY= # get from https://api.search.brave.com/
cli mcp list # list configured servers
cli mcp test my_server # verify connection
cli agents langchain -p my_agent "search for: Python MCP"
| Transport | Use when | Notes |
|---|---|---|
stdio | local process or Docker | default, most reliable |
sse | remote/HTTP server | requires running server URL |
For sse transport:
my_server:
transport: sse
url: http://localhost:8080/sse
| Concern | Path |
|---|---|
| MCP config | config/mcp_servers.yaml |
| MCP CLI commands | genai_tk/cli/commands_mcp.py |
| MCP docs | docs/mcp-servers.md |
| Example server | examples/mcp_server/server.py |
| Agent profiles | config/agents/langchain.yaml |