knowledge-base

Semantic search in a knowledge base lets you query indexed documents using natural language to retrieve existing context. Query the knowledge base before external research to avoid duplication.

Ingest notes and documents into shared or private collections using kb-ingest commands. Save substantive research results into memory files for future agents to query.

Yes, you can manage the knowledge base lifecycle with kb-setup and kb-collections commands to keep org data organized into distinct shared or private collections.

Querying a knowledge base before external research retrieves existing context from memory files and research notes, preventing duplication and ensuring you build upon previously indexed org knowledge.

Yes, ingest substantive research results into shared or private memory collections after completing research. Future agents can then query this knowledge base to retrieve the saved context.