Search knowledge bases semantically with vector embeddings and keyword fallback.
npx skills add https://github.com/BEKO2210/Firstbrain --skill search-beko2210
Please help me install this Agent Skill. Skill: search Source: https://github.com/BEKO2210/Firstbrain/tree/main/.agents/skills/search Command: npx skills add https://github.com/BEKO2210/Firstbrain --skill search-beko2210
This Skill helps you quickly find relevant information within your notes, even if you don't remember the exact keywords used, by understanding the meaning behind your search queries.
Search your vault for notes related to "effective time management techniques".
High-intent search queries and answers about installing and using this skill.
Semantic search finds notes by matching the conceptual meaning of your query rather than exact keywords. It uses vector embeddings to understand context, returning relevant information even if you don't remember the specific terms used.
Semantic search works by generating vector embeddings locally via @huggingface/transformers and storing them in an SQLite database. This allows the system to efficiently query and match notes based on conceptual relationships.
If AI search is unavailable, information retrieval falls back to a traditional keyword and tag matching mechanism. This ensures you can still find relevant notes, accompanied by generated excerpts showing their relevance.
Yes, @huggingface/transformers is required to perform semantic searches on your knowledge base. It handles the local vector embedding generation necessary to match query meaning and store the results in SQLite.
The best way to find notes about a vaguely remembered concept is using semantic search. It matches the meaning of your query, so searching for 'effective time management' will successfully retrieve notes written about 'optimizing workflows'.
The limitation of using keyword fallback for retrieving notes is that it relies on exact keyword and tag matching rather than conceptual meaning. This means it may miss relevant notes if the query terms differ from the original text.
XSpoonAiIndex local notes with vector embeddings for offline semantic search.
prathamchopra001Extract structured research findings with claims, confidence scores, and source details from academic paper excerpts.
ArogyaReddyCombine vector similarity and keyword matching with RRF, Linear, Cross-encoder, or Cascade fusion.
adeboldPerform vector-based semantic search against an AgentDB index.
