skills.rest
BermudaLocals
BermudaLocalsCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

qmd

Search local knowledge bases with hybrid BM25, vector, and LLM reranking.

1|1|Updated Apr 26, 2026
Search Tags:#mcp#notes#knowledge-base#vector-search#bm25#local-search#qmd
One-click install
npx skills add https://github.com/BermudaLocals/hermes-agent-lite --skill qmd-bermudalocals
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: qmd
Source: https://github.com/BermudaLocals/hermes-agent-lite/tree/main/optional-skills/research/qmd
Command: npx skills add https://github.com/BermudaLocals/hermes-agent-lite --skill qmd-bermudalocals

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Search local knowledge bases efficiently by combining full-text search, vector similarity, and LLM-powered reranking to retrieve relevant notes, transcripts, and documents without cloud dependency.

Core Features & Use Cases

  • Hybrid retrieval: BM25 keyword search plus vector search with LLM reranking for high-quality results.
  • Local-first knowledge management: index and query notes, meetings transcripts, and docs stored on-device.
  • MCP integration: expose Hermes Agent tools for automated workflows and automation.

Quick Start

Run a hybrid search across your local collections for a concept.

Frequently Asked Questions about qmd

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

FAQPage Schema
How do I search local notes and transcripts without sending data to the cloud?▼

Local knowledge search combines BM25 keyword search, vector similarity, and LLM-powered reranking to retrieve notes and transcripts without cloud dependency. It indexes text files stored on-device for private, offline document retrieval.

What is hybrid retrieval and how does it rank local search results?▼

Hybrid retrieval combines BM25 full-text keyword matching with vector similarity search to find relevant documents. An LLM reranker then refines these combined results to ensure high-quality, contextually relevant document retrieval from local knowledge bases.

Do I need Node.js and SQLite to run local vector search on my knowledge base?▼

Yes, running local vector search requires Node.js >= 22 and SQLite with extensions. The system also requires local model downloads to process your knowledge base and perform on-device LLM reranking for accurate hybrid retrieval.

Can I integrate local knowledge search with an automated agent workflow?▼

Yes, you can integrate local knowledge search into automated workflows using optional MCP integration. This exposes search tools to the Hermes Agent, enabling automated retrieval and querying of your on-device knowledge base.

What's the best way to find a specific concept across personal documentation files?▼

The best way to find a concept across personal documentation is running a hybrid search across your local collections. This combines BM25 and vector search with LLM reranking to retrieve highly relevant text files stored on-device.

Related Skills

Explore More Data & Analytics Skills→
GregsGreyCodeGregsGreyCode
6

qmd

Search local Markdown notes and transcripts with hybrid BM25, vector, and LLM reranking.

Community
Advanced
jason660519jason660519

qmd

Search local notes and transcripts with hybrid BM25, vector, and reranking retrieval.

Community
Intermediate
steipetesteipete
385.7k

qmd

Index local files and search them with BM25, vector, and hybrid queries.

Community
Advanced
acarmonagacarmonag
2

AgentDB Vector Search

Implement semantic vector search with AgentDB using HNSW indexing and hybrid search.

Community
Advanced