enable-semantic-search

Semantic search uses local AI embedding models to find information based on meaning and concepts rather than exact keywords. This allows queries like "PLG" to automatically locate notes containing "product-led growth" without matching phrases.

To enable semantic search, run the enable-semantic-search skill to automatically download and configure embedding, reranking, and query expansion models locally, setting up meaning-based search indexes for your vault.

Yes, local semantic search requires downloading and running local AI models for embedding, reranking, and query expansion. It also depends on bun, sqlite, and qmd, integrating with MCP servers for use within AI agents.

Yes, semantic search features smart collection discovery that intelligently analyzes vault content to automatically identify and create specialized search indexes for relevant sections like 'People' or 'Meetings'.

Semantic search integrates with existing skills like daily-plan and meeting-prep by providing contextually relevant information from your vault, enhancing these workflows with richer, meaning-based context retrieval.

Local AI semantic search requires sufficient computational resources to run embedding and reranking models locally. It also depends on a specific environment setup including bun, sqlite, and qmd, and integrates via MCP servers for AI agent compatibility.