semantic-search

Semantic search lets you find Obsidian notes by meaning instead of exact keywords. It applies embedding similarity to retrieve the most relevant matches, helping you surface connected ideas and answer questions over your personal knowledge base.

Yes, you can use LanceDB to index and retrieve your personal Zettelkasten notes. The Skill requires local note indexing with embeddings stored in LanceDB, then reads the highest-ranked notes to summarize the most useful findings for your query.

The best way to find related concepts is semantic note search. It re-indexes vault notes and retrieves the most relevant matches by embedding similarity rather than exact terms, surfacing prior research and connected ideas for your new question.

Yes, semantic search over personal notes requires local note indexing with embeddings and LanceDB storage. You must generate embeddings for your notes and store them in LanceDB before retrieving the highest-ranked matches for summarization.

Exact keyword search misses related ideas because it relies on matching specific terms rather than meaning. Semantic search solves this by applying embedding similarity to retrieve notes that share semantic meaning, even without overlapping keywords.

After retrieving notes from your Obsidian vault, the Skill reads the highest-ranked matches and summarizes the most useful findings. You simply ask the assistant to search your vault for notes about a topic and summarize the top matches.