qmd

Search markdown knowledge bases with hybrid lex, vector, and HYDE retrieval.

1|2|Updated Apr 13, 2026

Search Tags:#markdown #search #knowledge base #mcp #bm25 #semantic retrieval

Ready to run

npx skills add https://github.com/jgwill/dotagents --skill qmd-jgwill

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: qmd
Source: https://github.com/jgwill/dotagents/tree/main/skills/qmd
Command: npx skills add https://github.com/jgwill/dotagents --skill qmd-jgwill

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It helps you quickly find relevant information across markdown notes and documentation when you do not know where the answer lives.

Core Features & Use Cases

Lex/Vector/HYDE Searching: Combine keyword matching, semantic search, and hypothetical “answer-like” retrieval to improve recall.
Collection Filtering: Search within specific named collections (or across all collections when omitted).
MCP and HTTP Access: Query via an MCP server or a local HTTP API for flexible integration.
Use Case: When you need the exact doc section describing a feature (e.g., “rate limiter burst traffic”), run a hybrid query to retrieve the most relevant snippets from your knowledge base.

Quick Start

Ask qmd to search your markdown knowledge base for “how does the rate limiter handle burst traffic” and return the top 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 markdown documentation for specific technical phrases?▼

To search markdown notes, use a hybrid query combining BM25 lex matching, vector semantic search, and HYDE retrieval to find relevant snippets and documents across your knowledge base collections.

How does semantic retrieval improve knowledge base discovery?▼

Semantic retrieval improves knowledge base discovery by applying vector search and HYDE hypothetical answer generation to find relevant markdown documents even when exact keywords are missing or query intent is ambiguous.

Can I filter markdown searches within specific named collections?▼

You can filter markdown searches within specific named collections to narrow document discovery, or omit the filter to search across all configured collections in your knowledge base.

Do I need an MCP server to query my markdown notes?▼

An MCP server is not strictly required to query markdown notes; the system also supports a local HTTP mode or the qmd CLI for flexible integration and document retrieval.

What is the best way to find where a feature is described in markdown notes?▼

The best way to find where a feature is described is running a hybrid search query that combines keyword matching and semantic retrieval to return the most relevant doc sections from your notes.