Agent Hub MCP

description = "Register current agent with the MCP Agent Hub" prompt = """ Register the current agent with the MCP Agent Hub for multi-agent collaboration. I'll register this agent with the hub using the current project context. If you provide an agent ID as an argument, I'll use that, otherwise I'll auto-detect from the current directory. ⏺ agent-hub - register_agent (MCP)( id: {{args}} || basename of current directory, projectPath: current working directory, role: inferred from project type and context ) After successful registration, I'll show: - ✅ **Agent Profile**: Your registered ID, role, and detected capabilities - 📊 **Project Analysis**: Automatically detected technologies and skills - 🎯 **Registration Status**: Confirmation and next steps ## Hub Overview Then I'll fetch and display the current agent ecosystem: ⏺ agent-hub - get_hub_status (MCP)() This will show you: - 🤝 **Other Active Agents**: Who else is available for collaboration - ⚙️ **Capabilities Overview**: What skills are available in the hub - 🔗 **Collaboration Opportunities**: Potential matches based on complementary skills - 📡 **Hub Statistics**: Total agents, activity levels, recent registrations ## What's Next? After registration, you can: - Use `/hub:status` to see hub activity, all agents, and collaboration opportunities - Use `/hub:sync` to get your messages and workload ## Feature-Based Collaboration (Advanced) After registration, inform the user that the Agent Hub supports **feature-based collaboration** for complex multi-repository projects. **Explain when features are useful:** - When work spans multiple projects/repositories requiring coordination - Example: "Adding a new modality to an AI agent requires changes in the API (backend), UI (frontend), and agent support for modality (livekit agents)" - Any scenario where multiple agents need to coordinate changes and track dependencies **How to present the feature system:** - Features organize complex work into tasks with agent delegations - Each agent gets specific scope and creates subtasks for their part - Agents coordinate by updating progress and sharing context within the feature - The system provides structure for tracking who's working on what **When to suggest using features:** - If the user mentions work that clearly involves multiple repositories - When coordination with other agents becomes necessary - Let the user know features can be created when multi-agent coordination is needed If registration fails, I'll provide troubleshooting steps and check if the agent-hub MCP server is properly configured. """