Parse and index Markdown or Godot XML documentation into a SQLite database.
npx skills add https://github.com/emeraldwalk/skills --skill docs-reading
Please help me install this Agent Skill. Skill: docs-reading Source: https://github.com/emeraldwalk/skills/tree/main/docs-reading Command: npx skills add https://github.com/emeraldwalk/skills --skill docs-reading
This Skill simplifies the process of reading and referencing large technical documentation corpora, enabling fast and comprehensive information retrieval.
To start using the docs-reading skill, first, parse your documentation into the database using the following command:
uv run scripts/parse_docs.py --input /path/to/docs --db ./docs.db --corpus-name mylib --corpus-version 1.0
Once the documentation is parsed, you can use the agent-docs-search tool to search or retrieve information from the database:
agent-docs-search --db /path/to/skill/dbs/name.db search "your query" --corpus mylib --version 1.0
High-intent search queries and answers about installing and using this skill.
To parse and index technical documentation, run the parse_docs.py script via uv to process Markdown or Godot XML files, storing the corpus and version metadata in a local SQLite database for fast full-text search retrieval.
Yes, you can perform semantic search on Markdown documentation by generating optional sentence-embedding vectors using sentence-transformers, which are stored in the SQLite database to find semantically similar information.
Yes, you need Python and uv installed to set up a local documentation database, as uv manages the execution of parsing scripts and dependencies like numpy and sentence-transformers required for embedding generation.
The best way to search a complex library's documentation corpus is to index it into a SQLite database and use the agent-docs-search CLI tool, enabling quick full-text and semantic queries to retrieve only relevant sections.
Yes, this semantic search approach works with Godot XML documentation, as the parser supports both Markdown and Godot XML formats, indexing them into the SQLite database for comprehensive full-text and semantic querying.
Cissou34730Standardize Python dependency management with uv add, remove, sync, and run.
alsk1992Search local Markdown notes with BM25, vector, and hybrid modes.
zhuangbiaoweiManage an open-source embedding database for semantic search and retrieval.
JuanMS20Index Markdown and text files for local hybrid search with BM25, vectors, and reranking.