Search local Markdown collections with hybrid keyword and vector retrieval.
npx skills add https://github.com/overviewlabs/WHOX --skill qmd-overviewlabs
Please help me install this Agent Skill. Skill: qmd Source: https://github.com/overviewlabs/WHOX/tree/main/optional-skills/research/qmd Command: npx skills add https://github.com/overviewlabs/WHOX --skill qmd-overviewlabs
Local, on-device search across personal knowledge bases, notes, docs, and meeting transcripts to avoid cloud dependencies and speed retrieval.
Install qmd globally, add your document collections with qmd collection add, and start searching with qmd search or qmd query.
High-intent search queries and answers about installing and using this skill.
Local search of Markdown notes and docs is achieved through on-device indexing and retrieval, enabling fast keyword search without relying on cloud services. It supports personal knowledge bases and documentation workflows privately.
Hybrid search combines BM25 keyword matching with vector search to retrieve information from local knowledge bases. It then applies LLM reranking to the results to provide semantically accurate and contextually relevant outputs.
Yes, MCP integration exposes local search tools to your agent via stdio or an HTTP daemon. This setup allows native tool access for retrieving meeting transcripts and notes directly within your agent workflows.
Local vector search works offline for meeting transcripts and documentation by managing on-device embeddings. This approach ensures privacy and allows fast retrieval of contextual metadata without requiring an internet connection.
The best way to manage document collections for on-device search is by grouping Markdown files and adding them via collection commands. This contextual metadata organization directly improves the accuracy of hybrid search reranking.
cdeistopenedIndex and search local Markdown collections with BM25 and vector embeddings.
BEIRUXCall and manage MCP servers and tools via HTTP or stdio.
jason660519Search local notes and transcripts with hybrid BM25, vector, and reranking retrieval.
jhl-labsRetrieve relevant documents and augment LLM prompts with vector search.
