Index and search local knowledge bases with hybrid retrieval.
npx skills add https://github.com/jacardl/New-Radar --skill qmd-jacardl
Please help me install this Agent Skill. Skill: qmd Source: https://github.com/jacardl/New-Radar/tree/main/backend/frameworks/hermes-agent/optional-skills/research/qmd Command: npx skills add https://github.com/jacardl/New-Radar --skill qmd-jacardl
Personal knowledge management often requires fast, local search across notes, files, and transcripts without relying on cloud services.
Install qmd locally, index your collections, and run a search to retrieve relevant notes.
High-intent search queries and answers about installing and using this skill.
Local knowledge search lets you index and retrieve markdown notes, transcripts, and documents entirely on-device. The qmd Skill uses a hybrid retrieval engine combining BM25, vector search, and LLM reranking to find relevant content.
Hybrid retrieval combines BM25 keyword matching, vector embeddings, and LLM reranking to improve search quality. This approach ensures that local knowledge base searches return highly relevant notes and documents by blending lexical and semantic relevance.
Yes, local knowledge search with this Skill requires Node.js version 22 or higher and locally installed models. These prerequisites enable the on-device hybrid retrieval engine and MCP integration to function without relying on external cloud APIs.
Yes, you can expose local search tools to Hermes Agent through MCP integration. The Skill provides both a CLI and MCP interface, allowing an LLM agent to directly query your indexed local knowledge base and retrieve relevant documents.
Indexing local documents for LLM retrieval involves installing the qmd engine locally, pointing it at your collections of markdown notes and transcripts, and running the indexing process to enable fast, on-device hybrid search.
Local knowledge base search requires sufficient hardware to run locally installed models for vector embeddings and LLM reranking. Performance depends on your machine capabilities, and the system operates entirely offline without accessing cloud-based search APIs.
GregsGreyCodeSearch local Markdown notes and transcripts with hybrid BM25, vector, and LLM reranking.
thecloudtipsSynthesize sourced answers with inline citations and confidence ratings.
loongclaw-aiConnects Hermes Agent to external MCP servers via stdio or HTTP and registers their tools with prefixed names.
davpatel605-beepStore and retrieve vector embeddings for semantic search and document retrieval.
