Search local notes and transcripts with BM25, vector search, and LLM reranking on-device.
npx skills add https://github.com/matlee0409/cronus --skill qmd-matlee0409
Please help me install this Agent Skill. Skill: qmd Source: https://github.com/matlee0409/cronus/tree/main/optional-skills/research/qmd Command: npx skills add https://github.com/matlee0409/cronus --skill qmd-matlee0409
Local, on-device search for personal knowledge bases across markdown notes, meeting transcripts, and documentation with a hybrid retrieval pipeline and LLM reranking, all without cloud dependencies.
Search my notes for 'onboarding' using qmd.
High-intent search queries and answers about installing and using this skill.
Local, on-device search for markdown notes and documents uses a hybrid retrieval pipeline combining BM25, vector search, and LLM reranking. This retrieves personal knowledge bases directly on your machine without cloud dependencies.
On-device knowledge base search requires Node.js version 22 or higher and SQLite with extension support on macOS. You also need to download initial local GGUF models to enable the vector search and LLM reranking pipeline.
Yes, hybrid retrieval combines BM25 and vector search for local document retrieval. The pipeline then applies LLM reranking to deliver context-aware results across multiple indexed collections of notes and project files.
Local search tools integrate with Cronus Agent through exposed MCP tools and configurations. This allows the agent to directly access your indexed local knowledge bases and retrieve relevant notes or documentation.
Yes, local document search supports indexing meeting transcripts and project files. The system indexes these local files to enable fast multi-collection queries and context-aware retrieval across your personal knowledge management system.
JuanMS20Index Markdown and text files for local hybrid search with BM25, vectors, and reranking.
overviewlabsSearch local Markdown collections with hybrid keyword and vector retrieval.
GregsGreyCodeSearch local Markdown notes and transcripts with hybrid BM25, vector, and LLM reranking.
Chris-Chai-MinjaeIndex local text files for hybrid keyword and semantic search.
