secondbrain-search

To perform semantic search across your knowledge base, provide a natural language query to locate relevant notes, ADRs, and documentation using meaning-based similarity and full-text indexing. Results are returned as structured JSON with metadata for easier triage.

Filter knowledge base search results by applying entity, recent, from/to date, status, tag, and limit parameters to your semantic query. This narrows the retrieved notes and ADRs to match specific criteria within your project's indexed documentation.

Yes, you need qmd installed and initialized in your project to run semantic search on your notes. The search also requires access to the .claude/search index to produce structured JSON output from your knowledge base.

Yes, you can search for tasks and documentation across multiple projects using semantic similarity. The search retrieves relevant ADRs, notes, and documents, enriching the results with metadata from the project's knowledge store for easier triage.

Your knowledge base search returns structured JSON because the underlying qmd query produces structured output to enrich results with project metadata. This format facilitates easier triage and automated follow-up actions on the retrieved ADRs and notes.