# duckdb-vectors

> A frugal, embeddable semantic-search building block for agentic systems — hybrid vector + exact-token search over a local DuckDB file, no external vector database.

## Documentation

- [User guide](https://bernard357.github.io/duckdb-vectors/duckdb-vectors-guide): the straightforward developer workflow — index, search, integrate.
- [Technical design](https://bernard357.github.io/duckdb-vectors/duckdb-vectors-primitive): the architecture and building blocks of the package.
- [How DuckDB is used](https://bernard357.github.io/duckdb-vectors/duckdb-workbook): the DuckDB schema, hybrid ranking, and RRF fusion.

## Use cases

- [Set up a RAG agent](https://bernard357.github.io/duckdb-vectors/setup-rag-procedure): turn a markdown folder into a RAG retrieval layer.
- [Turn an LLM wiki into RAG](https://bernard357.github.io/duckdb-vectors/setup-llm-wiki-to-rag-procedure): chat over an incrementally managed wiki.
- [Search over skills](https://bernard357.github.io/duckdb-vectors/setup-skills-search-procedure): select agent skills by semantic search.

## Agents and MCP

- [Docker Agent guide](https://bernard357.github.io/duckdb-vectors/docker-agent-guide): wire an agent to the MCP-over-stdio `search` and `report` tools.
- [AGENTS.md](https://github.com/bernard357/duckdb-vectors/blob/main/AGENTS.md): how an agent adopts duckdb-vectors — install, SDK entry point, MCP tool.

## Resources

- [PyPI project](https://pypi.org/project/duckdb-vectors/): install with `pip install duckdb-vectors`.
- [Source repository](https://github.com/bernard357/duckdb-vectors): code, examples, and issues.
