doc-lookup

What problem does it solve?

Sifting through large documentation files to find specific information is time-consuming and inefficient, especially for AI agents that incur token costs for large context windows. This skill provides a targeted way to retrieve only relevant documentation excerpts, drastically improving efficiency and reducing cognitive load.

Core Features & Use Cases

• Anchor-Based Retrieval: Quickly locate and return specific documentation sections using explicit heading anchors (e.g., #cli-flags), ensuring the most precise results.
• Keyword & Heading Search: Find relevant sections even if the exact anchor is unknown, using fuzzy matching on heading text and content keywords for flexible discovery.
• Minimal Context Return: Returns only the top 1-3 most relevant sections with configurable line limits, drastically reducing token usage and providing focused context compared to loading entire documents.
• Use Case: When a router skill needs to understand "handler patterns" for a new node, it can use doc-lookup to retrieve just that specific section from docs/agents/package-maintainer.md, providing immediate, focused context without overwhelming the agent with irrelevant information.

Quick Start

Example: Find CLI Flag Documentation

Look up CLI commands section

python .claude/skills/doc-lookup/scripts/section_search.py
--query "cli-commands"
--paths docs/agents/backend-development.md
--top 1