ragflow-kb

What This Skill Does

The ragflow-kb skill serves as an intelligent bridge between OpenClaw AI agents and your local RAGFlow knowledge base. It is designed to interpret complex technical queries—ranging from Docker/Kubernetes container orchestration and system troubleshooting to internal operational guidelines—and retrieve precise, context-aware answers directly from your documents. By leveraging the RAGFlow API, the agent doesn't just provide generic information; it synthesizes answers based on your private repository, cites sources, and provides actionable, step-by-step instructions for resolving technical issues. The skill utilizes a streaming SSE architecture, ensuring real-time feedback during the generation process, which allows for complex, multi-stage reasoning and expert-level advice.

Installation

You can integrate this skill into your OpenClaw environment by running the following command in your terminal: clawhub install openclaw/skills/skills/chenkun-nuaa/ck-rag-skill This command automatically configures the necessary script dependencies located in /home/onestack/.openclaw/workspace/ragflow-kb/. Ensure your network has access to the RAGFlow API endpoint at http://172.28.20.46:30001/v1/conversation/completion.

Use Cases

• Container Lifecycle Management: Quickly identify why a container has entered an error state and execute recommended remediation steps.
• Technical Documentation Search: Instant retrieval of internal system configurations or deployment procedures without manually parsing through documentation folders.
• Operational Troubleshooting: Use the agent to diagnose server health issues by querying logs against known error patterns stored in the knowledge base.
• Infrastructure Guidance: Get summarized explanations of network modes, volume mounting, or environment variable configurations for your specific development environment.

Example Prompts

1. "The web-server container is stuck in a crash loop, can you check the logs and suggest a fix?"
2. "What is the recommended procedure for updating the Nginx configuration in our production environment?"
3. "Explain the differences between bridge and host network modes based on our internal infrastructure docs."

Tips & Limitations

• Patience is Key: The API uses a streaming output. Please wait for the full response to finish (typically 5-10 seconds) before interacting further.
• Use Verbose Mode: If you suspect an error, run the query with the -v flag to view detailed request/response headers and latency data.
• Contextual Awareness: The system reuses a conversation_id. If you want a fresh context, restart the agent session.
• Error Handling: If the skill reports a failure, verify your connection to the backend server. The skill is optimized with a 60-second max timeout, so complex queries exceeding this may need to be broken down.