MAMA vector database — v1.0.0
This release ships the prebuilt ChromaDB vector index that powers MAMA's
semantic-search tools (search_regulations, get_article). It is distributed
here rather than in git because of its size (~1.4 GB compressed).
Contents
mama-vectordb-chroma_v2.zip— 48,775 embedded chunks (~1,700 EU regulatory
documents) built withall-MiniLM-L6-v2.
Install
# from the repository root, after cloning + running setup_mama.py
unzip mama-vectordb-chroma_v2.zip -d data/
# → data/chroma_v2/ (loaded automatically by mcp_server.py)The other three MCP tools (list_acts, get_act_metadata,
get_legal_relationships) work without this download — they read the curated
metadata committed in data/.
See the README for full setup.