anydocs

What This Skill Does

anydocs is a versatile OpenClaw skill designed to transform fragmented, web-based documentation into a high-performance, locally searchable database. It functions as a bridge between your terminal and the vast ocean of online knowledge. By crawling websites, utilizing XML sitemaps, and optionally rendering JavaScript-heavy Single Page Applications (SPAs) via Playwright, anydocs creates a structured index of your preferred technical documentation. It supports advanced search modalities, including BM25-style keyword matching and regex pattern matching, enabling you to find specific function signatures or conceptual explanations without manual browser navigation.

Installation

To install the skill, execute the command: clawhub install openclaw/skills/skills/pektech/anydocs. Once installed, navigate to the skill directory and run pip install -r requirements.txt. For advanced documentation sites that rely on client-side JavaScript for content rendering, you must additionally install Playwright: pip install playwright==1.40.0 followed by playwright install. This ensures the crawler can fully extract information from modern web frameworks. After installation, configuration is handled via profiles, where you define the target URL and sitemap location using the anydocs.py config command.

Use Cases

• Offline Productivity: Search internal project documentation or API references while on a plane or in restricted network environments using cached local indexes.
• Agent Context Injection: Enable your AI agents to reference current library documentation automatically by searching the index built by anydocs before generating code.
• Technical Synthesis: Quickly compare syntax or patterns across multiple different frameworks or libraries by switching between pre-configured search profiles.
• API Exploration: Rapidly query specific endpoints or property definitions directly from your CLI, bypassing the need to hunt through navigation menus in a web browser.

Example Prompts

1. "anydocs, search the vuejs profile for the composition api pattern to update my state management logic."
2. "I need a quick reference for the discord API headers, can you use anydocs to fetch the documentation pages related to rate limiting?"
3. "Run a regex search in my internal-docs profile to find all instances where the deprecated 'auth_v1' function is still being referenced in our guides."

Tips & Limitations

• Rate Limiting: While the tool includes automatic rate limiting, index large documentation sites during off-peak hours to avoid triggering security blocks or exhausting server resources.
• Cache Management: Remember to periodically run cleanup commands to clear old cache files, especially if you index rapidly changing documentation sites. The default 7-day TTL is a good baseline, but adjust this via the profile configuration if your targeted docs update frequently.
• JavaScript Requirement: If a site appears empty after indexing, it is likely a client-side rendered app; ensure Playwright is installed to capture the DOM correctly.
• Storage: Large documentation sites with deep link trees can consume significant disk space; ensure your environment has adequate storage capacity before indexing extensive documentation portals.