qmd

Name: qmd
Availability: InStock
Author: drshailesh88

Index local documents and search them with BM25 and vector embeddings.

Updated Jan 8, 2026

Search Tags:#indexing #MCP #search #retrieval #vector #local #BM25

Ready to run

npx skills add https://github.com/drshailesh88/Krypto --skill qmd-drshailesh88

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: qmd
Source: https://github.com/drshailesh88/Krypto/tree/main/skills/qmd
Command: npx skills add https://github.com/drshailesh88/Krypto --skill qmd-drshailesh88

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Index and search local documents efficiently by indexing files and providing search over them using a hybrid of BM25 lexical scoring and vector similarity with optional reranking.

Core Features & Use Cases

Local indexing of documents, notes, and code repositories.
Hybrid search combining BM25 lexical ranking and vector embeddings with reranking.
Use cases include quickly locating relevant files in a project, recalling notes, and auditing documents.

Quick Start

Install qmd (Node) and start indexing:

Install: npm install -g qmd
Add a collection: qmd collection add /path/to/docs --name docs --mask "**/*.md"
Update index: qmd update
Search: qmd search "query"

Frequently Asked Questions about qmd

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

FAQPage Schema

How do I search local documents using both BM25 and vector embeddings?▼

You can search local documents by indexing files and querying them with a hybrid of BM25 lexical scoring and vector similarity. This approach blends keyword matching with semantic retrieval to locate relevant files in project collections.

What is the best way to index a local code repository for fast retrieval?▼

The best way to index a local code repository is using a CLI tool to scan files with specific masks. You add a collection pointing to your directory, update the index, and run keyword or vector searches to quickly locate relevant files.

Do I need Node installed to run a local BM25 search CLI?▼

Yes, you need Node installed because the CLI is installed via npm. Once installed globally, you can use it to index local documents, notes, and code repositories for hybrid search operations.

Can I use an MCP mode to blend lexical and semantic search results?▼

Yes, the CLI supports MCP mode to blend lexical and semantic results. This allows you to combine BM25 lexical ranking with vector similarity and optional reranking for improved retrieval accuracy.

How does hybrid search with reranking improve personal knowledge base retrieval?▼

Hybrid search with reranking improves personal knowledge base retrieval by combining BM25 keyword scoring with vector similarity. Optional reranking refines these blended results, ensuring fast and highly relevant document recall.