ClawdDocs: Your Expert Guide to Clawdbot's Own Documentation
11,600+ downloads and 118 stars — ClawdDocs is the most-starred documentation skill on ClawHub, and for good reason: it gives Clawdbot comprehensive, searchable, always-current access to its own documentation. When you're unsure how a feature works, need to find the right configuration option, or want to know what changed in the latest release, ClawdDocs is your first stop.
The Problem It Solves
Clawdbot is a powerful and rapidly evolving platform. Its documentation covers installation, configuration, skills architecture, MCP servers, cron jobs, sessions, the Gateway daemon, and dozens of other topics — spread across a large docs site that grows with each release.
Without ClawdDocs, you're context-switching: leaving Clawdbot, opening a browser, searching the docs, finding what you need, and returning. You may not find the right page. The documentation may have changed since you last read it. You can't ask follow-up questions.
With ClawdDocs, you never leave Clawdbot to look up Clawdbot.
The Documentation Stack
ClawdDocs is built on five scripts that form a complete documentation intelligence system:
sitemap.sh — Documentation Inventory
# Crawls the Clawdbot docs site and generates a structured index
./sitemap.shBuilds a map of all documentation pages, their URLs, titles, and sections. This is the foundation that the other scripts build on.
build-index.sh — Searchable Index
# Builds a full-text search index from the docs
./build-index.shProcesses all documentation pages into a searchable local index. The search is semantic enough to find relevant content from natural language queries, not just exact keyword matches.
search.sh — Query the Index
# Search the local documentation index
./search.sh "how do cron jobs work"
./search.sh "configure gateway daemon"
./search.sh "skill SKILL.MD format"Returns the most relevant documentation sections for your query. This is what powers the core "ask about docs" workflow.
fetch-doc.sh — Retrieve Full Pages
# Fetch a specific documentation page
./fetch-doc.sh /docs/skills/creating-skills
./fetch-doc.sh /docs/configuration/gatewayWhen you need the full content of a specific page — not just a search snippet — fetch-doc retrieves it completely. Useful for deep-reading a complex topic.
track-changes.sh — Documentation Changelog
# Check what's changed in the docs since last sync
./track-changes.shCompares the current docs site against the last indexed version and reports new pages, updated sections, and removed content. This keeps your ClawdDocs knowledge current with each Clawdbot release.
Usage
Getting Quick Answers
How do I set up a cron job in Clawdbot?
What's the format of a SKILL.MD file?
How does the isolated session mode work?
What environment variables does Clawdbot support?
Deep Dives
Show me the full documentation page on the Gateway daemon
What does the docs say about session types and their differences?
Find everything in the docs about skill versioning
Finding Recent Changes
What's changed in the Clawdbot docs recently?
Are there any new configuration options in the latest release?
What's new in the skills API since my last check?
Configuration Help
What are all the available configuration options for clawdbot config?
How do I configure custom system prompts?
What's the correct syntax for the .clawdrc file?
Decision Tree Navigation
ClawdDocs uses a decision tree navigation approach to handle ambiguous queries. Rather than returning a flat list of search results, it guides you through a structured path:
User: "How do I handle authentication?"
ClawdDocs: "Authentication in Clawdbot covers several areas:
1. MCP server authentication (API keys, OAuth)
2. Skill permissions and secrets management
3. 1Password / keychain integration
4. Gateway daemon authentication
Which area are you asking about?"
This interactive disambiguation improves answer quality significantly compared to naive keyword search, especially for topics that span multiple documentation sections.
Keeping the Index Fresh
As Clawdbot releases new versions, the documentation evolves. ClawdDocs handles this through:
- Manual refresh — ask "update the documentation index" to trigger a fresh crawl
- Change tracking — track-changes.sh highlights what's new so you don't have to re-read everything
- Version awareness — the skill notes which version of the docs it indexed, so you know if you're looking at current documentation
Update your documentation index — I just upgraded Clawdbot
What documentation changes came with the latest Clawdbot version?
ClawdDocs vs. Other Documentation Sources
| Approach | Setup | Freshness | Context retention | Follow-ups |
|---|---|---|---|---|
| ClawdDocs | Install once | Refresh on demand | Full | Yes |
| Browser search | None | Current | None | No |
| LLM knowledge | None | Outdated | None | Contextual |
Raw CLI --help | None | Always current | None | No |
The key advantage of ClawdDocs is the combination: searchable, contextual, conversational, and updatable.
Practical Tips
- Install early — ClawdDocs is most valuable from day one; the documentation learning curve for Clawdbot is steep, and having an in-context guide makes it much shallower
- Rebuild the index after upgrades — whenever you upgrade Clawdbot, ask ClawdDocs to refresh its index to pick up new documentation
- Use it for skills development — if you're writing a custom skill, ClawdDocs knows the complete SKILL.MD spec, the assets format, and the skill publishing workflow
- Ask about error messages — paste an error message and ask ClawdDocs if the docs mention it; often there's a troubleshooting section you'd otherwise miss
- Chain with execution — "show me how to configure X according to the docs, then help me configure it" combines documentation retrieval with hands-on assistance
Considerations
- Index staleness — if you don't refresh the index after upgrades, answers may reference outdated documentation; track-changes.sh helps detect this
- Coverage gaps — documentation that's been added very recently (within hours of a release) may not be indexed yet; if something seems missing, try refreshing the index
- Search quality — the local search index is good but not perfect for highly specialized queries; for obscure configuration options, fetch-doc on the specific page may be more reliable than search
- Documentation doesn't cover everything — some Clawdbot behaviors are only documented in release notes, GitHub issues, or community discussions; ClawdDocs covers the official docs site, not the full knowledge graph
The Bigger Picture
ClawdDocs is a meta-skill: a skill that helps you use all the other skills better by making Clawdbot's own documentation conversationally accessible. As Clawdbot becomes more complex — more configuration options, more skill APIs, more integration patterns — having an intelligent documentation guide becomes more valuable, not less. With 11,600+ downloads and the highest star count in its category, ClawdDocs has become essential infrastructure for anyone running Clawdbot seriously. The best way to learn a complex tool is to be able to ask it questions about itself.
View the skill on ClawHub: clawddocs