Search local wiki vaults with hybrid BM25/vector retrieval and grep fallback.
npx skills add https://github.com/RonanCodes/llm-wiki --skill search-ronancodes
Please help me install this Agent Skill. Skill: search Source: https://github.com/RonanCodes/llm-wiki/tree/main/.claude/skills/search Command: npx skills add https://github.com/RonanCodes/llm-wiki --skill search-ronancodes
Users need a reliable way to find relevant pages and passages inside a local wiki vault without re-deriving knowledge or manually scanning files, and this Skill provides ranked retrieval and sensible fallbacks.
Search the vault named my-research for "deployment patterns" and show the top results with snippets.
High-intent search queries and answers about installing and using this skill.
To search an Obsidian vault, you need ranked retrieval that scans local wiki pages to locate relevant documents and matching text snippets. This Skill uses qmd for hybrid BM25 and vector search to return ranked pages with snippets.
If qmd is not installed, the search falls back to recursive grep to locate matching text. It automatically extracts matching lines and frontmatter for context, and can offer Homebrew or Cargo installation for qmd.
Hybrid BM25 and vector search is a ranked retrieval method that combines keyword matching with semantic vector search and LLM re-ranking. It indexes local wiki pages on-device using qmd to return the most relevant pages and snippets.
Indexing an Obsidian-style vault for ranked search is handled automatically. The Skill detects whether a vault is indexed, performs initial indexing or updates, and parses qmd JSON output into readable results with matching snippets.
No external dependencies are strictly required. The tool detects missing qmd tooling and offers Homebrew or Cargo installation, but automatically falls back to recursive grep with frontmatter extraction if qmd is unavailable.
The best way to get ranked search results from a local vault is using qmd for hybrid BM25 and vector search with LLM re-ranking. It outputs JSON with relevant pages and matching text snippets, ideal for sourcing citations or triaging results.
JSai23Enable semantic search of local vault notes using qmd with BM25, vector search, and LLM re-ranking.
w-winterSearch indexed text corpora with semantic ranking and structured queries.
edonyzpcIdentify broken wikilinks, orphan notes, missing embeds, and inconsistent backlinks in Obsidian vaults.
paradox123Search markdown repositories with QMD ranked retrieval and rg fallback.
