comind

What This Skill Does

The comind skill integrates the OpenClaw AI agent directly into the CoMind人机协作平台. It serves as the primary bridge for AI members to handle task execution, markdown synchronization, interactive dialogue, and state management. By bridging OpenClaw with CoMind's infrastructure, this skill allows the agent to act as a formal team member, processing push tasks, scheduled inspections, and collaborative requests while ensuring that all communications and state updates follow the platform's standardized protocols.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/dqalex/comind. After installation, ensure you have configured your environment variables. You can set them in your systemd service file or .env file using COMIND_BASE_URL and COMIND_API_TOKEN. Alternatively, if you have an active WebSocket connection, the Gateway can dynamically request these credentials using the comind.config.request event, ensuring a seamless setup process without manual file editing.

Use Cases

This skill is ideal for teams utilizing CoMind for complex project management where AI needs to maintain consistency across distributed environments. Use cases include: automatically syncing markdown documentation updates when task status changes in CoMind; responding to real-time dialogue requests from team members via the chat channel; executing scheduled inspection tasks based on platform triggers; and maintaining a high-fidelity state panel that reflects the most current progress of long-running automated workflows.

Example Prompts

1. "Check for new tasks assigned to me on CoMind and initialize the state dashboard for the current project."
2. "Sync the latest documentation updates from the markdown repository to the CoMind knowledge base."
3. "Summarize the last 24 hours of task progress and post it to the team's collaborative chat channel."

Tips & Limitations

Always prioritize the MCP API channel for mission-critical operations, as it provides robust error handling, audit logging, and rate limiting compared to ephemeral chat channels. Be aware that the skill relies on a stable WebSocket connection for real-time responsiveness; if connectivity is lost, revert to HTTP-based polling via the MCP API. Ensure that all synced markdown files include proper Front Matter to prevent data ingestion errors. For security, never hardcode the COMIND_API_TOKEN in plain text inside your codebase; always utilize secure environment variable injection or the dynamic WebSocket configuration flow.