How does QMD work with Claude Code?

QMD includes an MCP (Model Context Protocol) mode that allows Claude to query your indexed local files as a tool, providing it with relevant context for your tasks.

Does QMD require an internet connection?

No, QMD performs indexing and searching locally on your machine, using Ollama for embeddings and reranking if configured.

Where is the search index stored?

By default, the QMD search index is stored locally on your machine under the `~/.cache/qmd` directory.

QMD is a local search and indexing CLI that uses hybrid search methods to help you find and retrieve information within your local files quickly.

What types of search does QMD support?

It supports BM25 (keyword-based), Vector (semantic/embedding-based), and Hybrid search which combines both methods for maximum accuracy.

QMD Local Search & Indexing

Name: QMD Local Search & Indexing
Author: ishuru

byishuru

生产力与工作流

Indexes local files and documentation to enable high-performance hybrid search via BM25 and vector embeddings.

关于

QMD is a specialized local search engine designed to integrate with Claude Code via MCP, allowing for lightning-fast retrieval of information from local file systems. It combines traditional BM25 keyword search with modern vector-based semantic search and reranking, powered by Ollama, to provide highly relevant context to your AI assistant. This skill is particularly useful for developers managing large internal documentation sets or codebases who need immediate, context-aware answers without leaving their CLI environment.

主要功能

Integration with Ollama for local embeddings and result reranking
MCP (Model Context Protocol) support for seamless integration with AI assistants
CLI-based document retrieval and management for local knowledge bases
Hybrid search combining BM25 keyword matching and vector embeddings
Local indexing of files and documentation with customizable masks
0 GitHub stars

使用场景

Providing semantic context to Claude from a private knowledge base using MCP
Quickly locating and reading specific snippets from indexed files via CLI commands
Searching through large local documentation folders for specific technical details

关于

主要功能

Integration with Ollama for local embeddings and result reranking
MCP (Model Context Protocol) support for seamless integration with AI assistants
CLI-based document retrieval and management for local knowledge bases
Hybrid search combining BM25 keyword matching and vector embeddings
Local indexing of files and documentation with customizable masks
0 GitHub stars

使用场景

Providing semantic context to Claude from a private knowledge base using MCP
Quickly locating and reading specific snippets from indexed files via CLI commands
Searching through large local documentation folders for specific technical details