mission-control

What This Skill Does

Mission Control is a sophisticated, Kanban-style task management engine designed specifically for OpenClaw AI agents. It transforms your workspace into an interactive task board where human-defined goals meet autonomous execution. By leveraging a centralized tasks.json file synced via GitHub, it creates a bidirectional communication channel between you and the AI. When you move a task to the "In Progress" column in the web dashboard, the agent automatically detects the state change, interprets the "Definition of Done" (DoD), and begins executing the work. It handles status transitions, subtask tracking, and status reporting, ensuring your workflow remains transparent and organized.

Installation

Installation is streamlined for rapid deployment. Simply initiate the process by commanding your agent: "Set up Mission Control for my workspace." The agent will verify your environment, ensuring the Tailscale CLI, Tailscale Funnel, and GitHub CLI are active. It then copies the necessary dashboard infrastructure to your repository, configures the ~/.clawdbot/mission-control.json file, and establishes the required webhooks on GitHub to bridge the UI to your local agent instance.

Use Cases

Mission Control is ideal for managing long-running projects, software development cycles, or complex multi-step workflows. Use it when you need to delegate tasks like "Research competitors," "Draft project proposal," or "Debug technical debt." It is particularly effective for teams or solo developers who require a visual board to track progress while offloading the manual execution of repetitive or defined tasks to an intelligent assistant.

Example Prompts

1. "Set up Mission Control for my workspace so I can start tracking my daily development tasks."
2. "Create a new task in the backlog called 'API Documentation' with a priority of high and add a subtask for generating the initial draft."
3. "Move the task 'Refactor Auth Module' to in-progress and let me know once you have finished the subtasks."

Tips & Limitations

To get the most out of Mission Control, ensure your "Definition of Done" is clearly written, as the agent relies on this to determine task completion. Always keep your local workspace synced with GitHub to avoid state conflicts. Note that this skill requires an active internet connection for the GitHub webhook bridge. Avoid moving multiple complex tasks to "In Progress" simultaneously to prevent agent context overload, and remember to perform regular manual reviews of tasks in the "Review" column to provide necessary feedback for the agent.