mqtt-agent-messenger

What This Skill Does

The mqtt-agent-messenger skill serves as the primary communication backbone for OpenClaw agents operating within a networked environment. By leveraging the lightweight MQTT protocol, this skill enables agents to broadcast their identities, roles, and real-time activity statuses. It acts as a pub/sub message broker interface that allows a decentralized fleet of agents to discover, coordinate, and monitor one another. When an agent performs an action or changes its focus, it publishes a status update; other agents can simultaneously subscribe to these updates to build a shared situational awareness map.

Installation

To integrate this communication capability into your OpenClaw environment, ensure you have the required Python dependencies installed. Run the following command in your terminal:

pip install paho-mqtt typer

Once the dependencies are met, install the skill directly from the OpenClaw repository using the hub command:

clawhub install openclaw/skills/skills/hyunwoo91/communication-mqtt

Verify your installation by locating your unique agent_id in the ~/.openclaw/openclaw.json configuration file, which will be used as the identifier for all MQTT messages.

Use Cases

• Multi-Agent Orchestration: Coordinate complex tasks where one agent (e.g., a researcher) needs to notify another (e.g., a coder) that a specific requirement has been met.
• Agent Health Monitoring: Track the operational status of long-running background agents to ensure they haven't crashed or entered an idle state unexpectedly.
• Shared Context Discovery: Allow new agents entering the network to immediately recognize the roles and current objectives of existing active agents.

Example Prompts

1. "Publish my current status as 'analyzing data' to the MQTT broker using the mqtt-agent-messenger skill."
2. "Subscribe to all intro messages from other agents for the next 30 seconds to see who is online."
3. "Check if agent-2 is still active by monitoring their status messages via MQTT."

Tips & Limitations

• Network Dependency: This skill relies on an external MQTT broker. Ensure that your network environment allows MQTT traffic, as strict firewalls may block these packets.
• Naming Conventions: Keep your agent_id consistent across your configuration to ensure other agents can accurately filter and target your messages.
• Latency: While MQTT is highly efficient, ensure that your broker is geographically or logically close to your agents to minimize latency during high-frequency status updates.
• Privacy: Be aware that all messages sent via this skill are broadcast to the broker. Avoid publishing sensitive data or secrets that you do not want visible to other agents on the same broker network.