clawctl

What This Skill Does

clawctl acts as the central nervous system for OpenClaw agent fleets. It provides a robust, standardized coordination layer that allows multiple AI agents to collaborate, manage tasks, and communicate efficiently. By providing a shared database-backed interface, it enables agents to synchronize their activity, track task lifecycles, and maintain visibility over the entire fleet's operations.

Installation

To install the skill, execute the following command in your terminal:

clawhub install openclaw/skills/skills/lludlow/clawctl

Once installed, you must initialize the storage layer using clawctl init. It is recommended to verify your configuration by running clawctl whoami to ensure your identity and database path are correctly set. You can customize the location of your task database by setting the CLAW_DB environment variable.

Use Cases

clawctl is ideal for scenarios where multiple autonomous agents must divide and conquer complex projects.

• Multi-Agent Orchestration: Agents can automatically hand off sub-tasks to each other, ensuring that no work is dropped during complex pipelines.
• Async Collaboration: By utilizing the built-in messaging system, agents can post status updates, block reports, or questions to a common thread, preventing communication silos.
• Task Tracking: It provides a structured Kanban-style board and task list, allowing human supervisors to see exactly what is in progress, pending, or under review at any given time.
• Operational Rhythm: Ensures consistent agent behavior by enforcing a standard workflow: checking in, clearing the inbox, claiming, and executing tasks.

Example Prompts

1. "Check the current fleet status and tell me which agents are currently working on a high-priority task."
2. "I am blocked on the documentation task. Use clawctl to mark task #402 as blocked by the research team and send an alert to them."
3. "List all my pending tasks and identify the one with the highest priority to start working on next."

Tips & Limitations

• Consistency is Key: The system functions best when all agents adhere to the checkin -> inbox -> next -> claim -> start -> done flow. Deviating from this can lead to data inconsistencies in the task database.
• Metadata usage: Make heavy use of the --meta flags when performing actions. This allows you to attach context such as file paths or specific artifact locations directly to the task, which is invaluable for long-running processes.
• Database Sensitivity: Since clawctl relies on a local SQLite database, ensure that the path defined in CLAW_DB is persistent and backed up if you are running in transient cloud environments.