qmd

Index and search private notes locally with hybrid retrieval.

1.2k|116|Updated Mar 19, 2026

Search Tags:#knowledge-base #vector-search #bm25 #local-search #qmd #hybrid-retrieval

Ready to run

npx skills add https://github.com/math-inc/OpenGauss --skill qmd-math-inc

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: qmd
Source: https://github.com/math-inc/OpenGauss/tree/main/optional-skills/research/qmd
Command: npx skills add https://github.com/math-inc/OpenGauss --skill qmd-math-inc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Locally indexes and searches private notes, docs, and transcripts to enable fast, private knowledge retrieval without cloud services.

Core Features & Use Cases

Hybrid retrieval combines BM25, vector search, and LLM reranking to deliver relevant results across personal collections.
Use Case: quickly locate a meeting summary or a keyword across weeks of notes and documents.

Quick Start

Run a local search query across your notes with qmd to retrieve relevant results.

Frequently Asked Questions about qmd

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I search private notes and documents locally without cloud services?▼

Local hybrid search indexes your private notes and documents using a retrieval engine that runs entirely on your machine, ensuring fast and private knowledge retrieval without cloud services.

What is hybrid retrieval and how does it work for local knowledge bases?▼

Hybrid retrieval combines BM25 keyword matching, vector search, and LLM reranking to deliver highly relevant search results across your local knowledge base, optimizing both exact keyword matches and semantic context.

Can I use vector search and BM25 for offline knowledge management?▼

Yes, you can use vector search and BM25 for offline knowledge management. The system indexes your data locally, allowing you to query collections securely without an internet connection.

Do I need embedding models or vector databases to run local document search?▼

Optional embedding or vector models are required to enable vector search capabilities. Local data storage is necessary to hold your notes, docs, and transcripts for the indexing process.

What is the best way to find a specific meeting transcript across weeks of notes?▼

The best way to find a specific meeting transcript is to use a local hybrid search query that scans indexed notes and documents, quickly locating summaries or keywords across large personal collections.

How do I access and query local search collections using CLI or MCP tooling?▼

You access and query local search collections by using CLI or MCP tooling integrated with the hybrid retrieval engine, enabling direct commands to search your indexed notes and transcripts.