Extend agent capabilities and tools
self-customizeskillsetup L3★29,740
nanocoai/nanoclaw ↗What it does
Query vector database over embedded context with relevance ranking
Best for
Finding relevant context in large knowledge bases when semantic search beats keyword/BM25 matching.
Inputs
- · query text
- · optional metadata filters
- · result limit
Outputs
- · ranked matching chunks with scores
- · chunk metadata
Requires
- · vector DB client
- · embedding model
Preconditions
- · chunks indexed with embeddings
- · database populated
Failure modes
- · poor embedding quality degrades result relevance
- · metadata filters exclude all results
- · query too short to embed meaningfully
Trust signals
- · vector similarity ranking
- · optional metadata filters
- · relevance scores per result