multi-agent-cn

What This Skill Does

The Multi-Agent Dispatcher (multi-agent-cn) transforms your OpenClaw agent from a single-task executor into a sophisticated team manager. Acting as a dedicated coordinator, the primary agent evaluates incoming tasks and delegates them to a team of five specialized persistent sub-agents (Alpha, Bravo, Charlie, Delta, Echo). By employing a strict protocol—where the dispatcher first acknowledges the task and then triggers specialized workers via 'sessions_spawn'—this skill ensures that tasks are handled with appropriate expertise. It supports complex multi-task decomposition, allowing users to trigger parallel workflows for non-dependent operations, ensuring your AI workspace remains organized and efficient.

Installation

To integrate this system into your workflow, execute the following command in your terminal: clawhub install openclaw/skills/skills/be1human/multi-agent-cn After installation, you should immediately review the internal configuration to customize the dispatcher's persona and the specialized roles of the sub-agents to fit your specific professional or creative requirements.

Use Cases

This skill is ideal for complex project management where tasks require different types of analysis or execution. Use it for:

• Software development cycles where code writing, review, and documentation can happen concurrently.
• Research projects involving simultaneous data gathering and analysis across multiple sources.
• Administrative workflows that require parsing disparate requests, such as filing reports while drafting follow-up emails.

Example Prompts

1. "I need to rebuild the landing page navigation and simultaneously check the documentation for the new API integration; please handle these."
2. "We have three bugs in the payment module: one in the validation logic, one in the database schema, and one in the UI. Can you coordinate the team to fix these?"
3. "Please conduct a market research study on current AI tools for developers, and while that's running, draft a summary of our team's current development roadblocks."

Tips & Limitations

To get the most out of this skill, ensure you follow the 'first response, then spawn' rule. The dispatcher agent must communicate its intent before invoking the tool, or the system may experience logic gaps. Remember that task delegation works best when requests are clearly separated into independent sub-tasks. Over-complicating dependencies can lead to inefficiencies; when tasks are sequential, allow the agent to finish the first step before initiating the next. Always assign a 'sessionKey' to avoid state confusion, and keep your primary agent strictly in the 'coordinator' role to prevent it from consuming local execution resources meant for the sub-agents.