Index and query markdown files with lexical and vector search.
npx skills add https://github.com/woosul/wikey --skill qmd-woosul
Please help me install this Agent Skill. Skill: qmd Source: https://github.com/woosul/wikey/tree/main/tools/qmd/skills/qmd Command: npx skills add https://github.com/woosul/wikey --skill qmd-woosul
qmd solves the problem of efficiently searching and retrieving information from markdown knowledge bases, notes, and documentation. It helps users find documents, look up information, and perform in-depth searches without manually scanning through files.
To search for the term "CAP theorem" in the "docs" collection, run: qmd query "CAP theorem consistency" --collections docs
High-intent search queries and answers about installing and using this skill.
Markdown search engines use lexical search for exact terms and vector search for natural language questions. This allows you to retrieve information from notes and documentation using either exact phrases or conceptual queries.
You can query a specific markdown collection by running a local CLI command with a collections flag. For example, use `qmd query "search terms" --collections docs` to filter results to the "docs" subset.
Yes, markdown search works with the Markdown Content Provider (MCP) protocol for server-side queries. This integration provides a seamless search experience for retrieving indexed markdown documents over a server connection.
Hypothetical document search (hyde) in markdown retrieval accepts answer passages as queries. This vector-based method matches the semantic meaning of the passage to find relevant information within indexed knowledge bases.
The limitation of lexical search for markdown knowledge bases is that it relies on matching exact terms or phrases. It may miss conceptually relevant documents if the exact vocabulary is not used, unlike vector-based search methods.
MrGKanevIndex Markdown notes locally and search via keyword, semantic, or hybrid modes.
rukikotooQuery the econometrics RAG knowledge base to retrieve relevant document chunks.
kingkillerySearch markdown collections with full-text and semantic queries.
SignmanalStore and query vector embeddings with metadata filtering for RAG applications.
