openclaw-doc-helper

# OpenClaw Documentation Helper This skill helps answer OpenClaw-related questions by automatically querying the official documentation at docs.openclaw.ai. ## When to Use Use this skill when the user asks about: - OpenClaw configuration and setup - Available commands and options - Troubleshooting issues - Architecture and design - Best practices - Any OpenClaw-specific functionality ## Workflow 1. **Identify the query topic** - Understand what specific OpenClaw topic the user is asking about 2. **Search/fetch relevant documentation** - Use `web_fetch` to retrieve documentation from docs.openclaw.ai: - For general questions: Start with `https://docs.openclaw.ai` - For specific topics: Try `https://docs.openclaw.ai/<topic>` (e.g., `/faq`, `/troubleshooting`, `/config`) - Use `web_search` with site:docs.openclaw.ai if unsure where to look 3. **Analyze the documentation** - Read and understand the fetched content 4. **Formulate a comprehensive answer** - Based on the official documentation, provide an accurate response 5. **Cite sources** - Mention that the information comes from docs.openclaw.ai ## Configuration Validation (User-Guided) When suggesting configuration changes, guide the user to validate and view their configuration: **Validate configuration:** ```bash # Validate entire configuration openclaw config validate ``` **View specific configuration sections:** ```bash # View all plugins configuration openclaw config get plugins # View gateway settings openclaw config get gateway # View models configuration openclaw config get models # View channels configuration openclaw config get channels # View agents configuration openclaw config get agents # View specific channel (e.g., Feishu) openclaw config get channels.feishu # View nested values openclaw config get agents.defaults.model.primary openclaw config get channels.whatsapp.enabled ``` **Explain the validation output to the user:** - If `valid: true` → configuration is correct, can proceed - If errors → explain what needs to be fixed based on the error messages - If warnings → inform user but note it's not blocking ## Key Documentation URLs - Main docs: https://docs.openclaw.ai - FAQ: https://docs.openclaw.ai/faq - Troubleshooting: https://docs.openclaw.ai/troubleshooting - Configuration: https://docs.openclaw.ai/config - GitHub: https://github.com/openclaw/openclaw ## Example Usage User: "How do I configure a new channel?" Action: 1. Fetch `https://docs.openclaw.ai/config` or search for "channel configuration" 2. Analyze the configuration documentation 3. Provide step-by-step instructions based on official docs 4. Remind user to run `openclaw config validate` after making changes ## Notes - Always prefer official documentation over general knowledge - If documentation is unclear or missing, say so and suggest checking the GitHub repo - Keep answers concise but complete - **Security Note**: This skill does NOT execute any local commands. Users must run validation commands themselves. ## Safety & Security This skill is **instruction-only** with no code execution: - ✅ Only uses `web_fetch` and `web_search` tools - ✅ Does NOT execute local commands or access files - ✅ Does NOT require API keys or credentials - ✅ Users retain full control over their system