MissionControl
Autonomous Multi-Agent Orchestration
Your AI squad, on autopilot. A Commander agent coordinates your OpenClaw agents — assigning tasks, routing work, and running standups. Agents check in every 15 minutes, pick up work, talk to each other, and report back. Like Bhanu Teja's Mission Control, but file-based with zero external dependencies.
What it does
Commander agent — central orchestrator for your squad
Kanban task board (backlog → assigned → in-progress → review → done)
Agent-to-agent comms with @mentions
Smart escalation (auto-reassign stuck tasks)
On-demand markdown dashboard
Skills (5)
Installed into your agent's workspace
mission-control-agentMissionControl agent check-in: read assigned tasks, DO the work, respond to comms, update task status, write standup
Trigger phrases
missioncontrol-dashboardGenerate a MissionControl markdown dashboard: kanban status, agent standups, blockers, and priorities.
missioncontrol-escalateEscalate blocked tasks: flag blockers, reassign work, notify user of critical risks
Trigger phrases
missioncontrol-orchestrateCreate, assign, and track tasks across agents using MissionControl file-based kanban
Trigger phrases
missioncontrol-registryManage the MissionControl registry.json roster: discover agents, set roles/capabilities, show squad status.
Automations (3)
MissionControl — Daily Standup
enabled09:00 AM · weekdays (UTC)
cron: 0 9 * * 1-5 · tz: Etc/UTC
What the agent does on this trigger
Daily standup. Read registry.json and standups/ for today. If agents are missing standups, write a comms message asking them to run mission-control-agent check-in. Then synthesize: wins, priorities, blockers, risks, and next actions. Save a markdown report into ~/.openclaw/mission-control/standups/YYYY-MM-DD__commander.md.
MissionControl — Weekly Retrospective
enabled05:00 PM · every Fri (UTC)
cron: 0 17 * * 5 · tz: Etc/UTC
What the agent does on this trigger
Weekly retro. Review tasks done vs planned (tasks/done, tasks/review, tasks/in-progress). Identify process bottlenecks, recurring blockers, and propose 3 improvements. Generate a concise retro report and update USER.md recommendations if needed.
MissionControl — main check-in
enabledat minutes 0,15,30,45 of every hour · every day (UTC)
cron: 0,15,30,45 * * * * · tz: Etc/UTC
What the agent does on this trigger
MissionControl check-in. Run the mission-control-agent skill: check assigned tasks in ~/.openclaw/mission-control/tasks/, read comms in comms/inbox/ addressed to you, pick up and work on any assigned tasks, update task status as you make progress, and write a standup to standups/ if needed.
Integrations & requirements
- Primary integration mode
- Local filesystem orchestration
- MCP status
- No dedicated MCP integration
- Supported today
- Shared task board under `~/.openclaw/mission-control/`
- Cron-driven check-ins and local agent-to-agent comms
- Planned / not yet implemented
- ○No external service layer or dashboard API implemented
Files this agent creates
Written under ~/.openclaw/workspace-missioncontrol/
YYYY-MM-DDTHHMMSSZ__from-<yourId>__to-<theirId>__topic.mdstandups/YYYY-MM-DD__<agentId>.mdstandups/YYYY-MM-DD__dashboard.mdregistry.jsonInstall & configure
1. Install the agent pack
tinkerclaw install missioncontrol
2. Configure your preferences
Edit the per-agent USER.md to set tone, timezone, and any agent-specific preferences:
~/.openclaw/workspace-missioncontrol/USER.md
3. Keep it up to date
tinkerclaw update missioncontrol
Highlights
- ✦Agents autonomously pick up tasks, do the work, and hand off
- ✦File-based orchestration — no external DB or API needed
- ✦Installs into all existing agents automatically