Rebuild the FAISS vector index for updated semantic search accuracy.
npx skills add https://github.com/Abilityai/cornelius --skill refresh-index
Please help me install this Agent Skill. Skill: refresh-index Source: https://github.com/Abilityai/cornelius/tree/main/.claude/skills/refresh-index Command: npx skills add https://github.com/Abilityai/cornelius --skill refresh-index
Ensures that your semantic search results accurately reflect the latest changes in your knowledge base by rebuilding the FAISS index.
Run the refresh index skill to rebuild the FAISS index.
High-intent search queries and answers about installing and using this skill.
You can rebuild the FAISS vector index to update semantic search accuracy for your knowledge base. This process re-indexes your local brain notes so that recent modifications and additions are discoverable through search queries.
Semantic search returns outdated results because the FAISS vector index has not been rebuilt to include recent content changes. You must rebuild the index to synchronize the search database with your latest brain notes.
You should rebuild the FAISS vector index whenever your knowledge base content is frequently updated. Running the rebuild process after adding or modifying notes ensures that semantic retrieval remains accurate and current.
Yes, you need bash and FAISS dependencies to rebuild your local search index. The rebuild process requires executing a bash script that processes and indexes your brain notes using the FAISS vector search library.
The best way to maintain search accuracy for frequently updated notes is to periodically rebuild the FAISS vector index. This index maintenance guarantees that your semantic search queries retrieve the latest insights from your knowledge base.
AbilityaiAnalyze recent notes against a knowledge base using local brain search.
EndUser123Store and semantically search project knowledge locally with a FAISS-backed index.
jchildreeSearch, create, and organize notes within an Obsidian vault.
sheawinklerSearch local markdown and text collections using hybrid BM25 and vector ranking.
