Search Knowledge Stores

You can search indexed library sources by executing a query to retrieve relevant code snippets and documentation using vector, full-text, or hybrid search modes. Results can be filtered by score, relevance, and quantity.

Vector search matches semantic meaning, full-text search matches exact keywords, and hybrid search combines both approaches. Searching indexed knowledge stores with these modes facilitates retrieving specific code implementations or relevant documentation snippets based on query intent.

Yes, searching indexed knowledge stores supports detailed result filtering by score, relevance, and quantity. This allows you to configure the number of results returned and set detail levels to retrieve specific code implementations or documentation snippets.

To find how a specific function is implemented across dependencies, search your indexed knowledge base using a query targeting that function. The search retrieves relevant code snippets and documentation from indexed library sources based on your specified relevance thresholds.

Yes, searching indexed library sources requires the mcp__bluera-knowledge__search tool for execution. This dependency enables the semantic, full-text, or hybrid search modes used to retrieve code implementations and documentation snippets from the knowledge stores.

Semantic search across indexed library sources is limited to previously indexed data, meaning unindexed dependencies cannot be queried. Additionally, filtering by score and relevance thresholds may exclude tangentially related code implementations or documentation snippets if set too restrictively.