A local search and indexing CLI utilizing BM25 and vector search with MCP support for AI agent integration.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install qmd
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install qmd using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is qmd?
qmd is a powerful command-line interface designed to index and search local files using a combination of BM25, vector search, and reranking. It allows developers to turn local directories into searchable knowledge bases, making it an essential component for those building advanced Openclaw Skills. By utilizing local LLM providers like Ollama, qmd provides privacy-focused, semantic retrieval capabilities directly on your machine.
Whether you are managing large technical documentation sets or navigating complex codebases, qmd provides the speed of keyword search with the intelligence of vector-based understanding. Its support for the Model Context Protocol (MCP) ensures it can be seamlessly integrated into modern AI agent workflows.
qmd Use Cases
- Creating searchable indices for large local markdown documentation sets
- Implementing semantic vector search across private code repositories
- Providing high-context data to AI agents via the MCP mode
- Fast retrieval of specific document segments for RAG (Retrieval-Augmented Generation) workflows
How qmd Works
- Users define a collection by pointing qmd to a directory and defining a file mask.
- The indexing process generates a local cache containing BM25 indices and vector embeddings.
- Semantic embeddings and reranking are handled locally through a connection to Ollama.
- Users perform queries using either traditional keyword matching, semantic vector search, or a hybrid of both.
- The tool can be run in MCP mode to allow AI agents to programmatically query the local index.
qmd Setup
To add this to your Openclaw Skills, install qmd using the following command:
npm install -g https://github.com/tobi/qmd
Ensure Ollama is running at the default URL or set your environment variable:
export OLLAMA_URL="http://localhost:11434"
Initialize and update your index:
qmd collection add ./docs --name my-docs --mask "**/*.md"
qmd update
qmd Data Schema & Taxonomy
qmd manages its data through a structured local cache. Below is how it organizes its search environment:
| Attribute | Description |
|---|---|
| Cache Location | Files are stored in ~/.cache/qmd by default. |
| Collections | Defined by a name, path, and file-matching mask. |
| Search Modes | Supports BM25 (keyword), Vector (semantic), and Hybrid. |
| Retrieval | Allows fetching specific lines using the syntax path:start_line -l length |
qmd Advanced Features
- Model Context Protocol (MCP) support for seamless AI agent integration
- Hybrid search functionality combining lexical and semantic relevance
- Local reranking of results using Ollama-hosted models
- Configurable file masking for granular control over indexed content
SKILL.md
Loading
Related Openclaw Skills
Peekaboo
A comprehensive macOS UI automation CLI designed to capture screens, inspect elements, and drive user inputs for AI agents.
OrderCLI Foodora & Deliveroo Assistant
A command-line interface for tracking active food deliveries, viewing order history, and performing reorders on Foodora and Deliveroo.
Oracle AI Context Bundler
A specialized CLI tool that bundles prompts and source files into a single context-rich request for advanced AI model reviews and debugging.
OpenHue
A powerful CLI integration for controlling Philips Hue smart lighting systems directly through your AI agent.
sag
A high-performance CLI for ElevenLabs text-to-speech that brings macOS-style simplicity to professional AI voice generation.
Slack Actions
A robust integration for managing Slack workspace interactions, including messaging, reactions, and channel organization via AI agents.
Featured*
