haiku-rag
Opinionated agentic RAG powered by LanceDB, Pydantic AI, and Docling
AI & MLpython
How to use
haiku-rag is an MCP server that lets your AI assistant call models, embeddings, or ML services as tools.
- Best for chaining an assistant into other AI services and pipelines.
- Works in any MCP-compatible client — Claude Desktop, Claude Code, Cursor, and others.
- Add it once to your client config; its tools then appear to the assistant automatically.
Install
Distributed as a PyPI package. Prerequisite: Python 3.10+ and uv — https://docs.astral.sh/uv/.
- 1Install the prerequisite above if you don't have it.
- 2Open your Claude Desktop config (create it if missing):
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- macOS:
- 3Add this under
mcpServers(merge if the key already exists):{ "mcpServers": { "haiku-rag": { "command": "uvx", "args": ["haiku-rag"] } } } - 4Save the file and fully quit, then reopen Claude Desktop.
- 5Click the tools (🔌) icon — haiku-rag's tools should now be listed.
Capabilities
No capabilities documented for this server yet.
Reviews (0)
No reviews yet. Be the first to review this server.
