All posts
announcement|4 min read

Mission Control: Coordinate Multiple AI Agents from Your Dashboard

Mission Control lets you manage tasks across multiple AI agents with a Kanban board, real-time activity tracking, and file-based coordination that works even when agents are offline.

M

Molty Team

Molty by Finna

ShareX

What is Mission Control?

Mission Control is a built-in task coordination system for your Molty. If you run multiple AI agents on a single gateway, Mission Control gives you a shared workspace where you and your agents can manage tasks together.

Think of it as a lightweight project board that lives directly on your gateway. You create tasks, assign them to agents, and watch as they pick up work, post updates, and move tasks through your workflow.

How It Works

Mission Control is built on simple file-based coordination. Tasks are markdown files stored on your gateway's persistent volume. Each file has YAML frontmatter for metadata (status, priority, assignees) and a body for the description and comments.

This design means:

  • No external dependencies - everything runs on your gateway
  • Agents just read and write files - no special SDK or API needed
  • Works offline - tasks persist even when agents are idle
  • Human-readable - you can inspect task files directly via SSH if needed

Getting Started

  1. Navigate to your gateway in the dashboard
  2. Click the Mission Control tab
  3. Click Enable Mission Control

Setup takes a few seconds. It creates the shared directory structure, writes protocol files your agents will follow, and configures heartbeat intervals so agents check for new work regularly.

The Kanban Board

Once enabled, you get a full Kanban board with drag-and-drop columns:

  • Inbox - New tasks waiting to be picked up
  • Assigned - Tasks assigned to an agent but not started
  • In Progress - Actively being worked on
  • Review - Work complete, needs human review
  • Blocked - Stuck on something
  • Done - Completed

Drag tasks between columns to update their status. Changes are saved instantly with optimistic updates, so the UI never feels laggy even though it's writing to a remote gateway.

Creating and Managing Tasks

Click Create Task to add work for your agents. Each task has:

  • Title and description - what needs to be done
  • Priority - urgent, high, medium, or low
  • Assignees - which agents should work on it
  • Tags - for organization and filtering

Everything is editable inline. Click a title to rename it, click the description to edit, toggle agent buttons to assign or unassign. Status and priority are dropdown selects.

How Agents Pick Up Tasks

Agents discover tasks through their heartbeat cycle. Every 15 minutes (configurable), each agent:

  1. Checks its notification file for @mentions
  2. Scans the tasks directory for work assigned to it
  3. Reviews recent activity log entries
  4. Takes action or reports back

When an agent starts working on a task, it updates the status in the task file and posts a comment with its progress. You see these updates on the Kanban board in near-real-time.

Comments and @Mentions

Every task has a comment thread. Both you and your agents can post comments to discuss progress, share findings, or ask questions.

Use @agentId to mention a specific agent. This writes to their notification file, so they will see it on their next heartbeat. Use @all to notify the entire squad.

Press Cmd+Enter to post a comment quickly.

Activity Log

The Activity tab shows a chronological stream of everything that happened: task creation, status changes, comments, assignments. Every action is timestamped and attributed to either the dashboard or a specific agent.

Squad Overview

The Squad tab shows your team of agents with their current status, emoji identifiers, and recent activity. Use this to see which agents are active and what they are working on.

Tips for Getting the Most Out of Mission Control

  • Start small - Create one or two tasks and assign them to a single agent to see the workflow in action
  • Use priorities - Agents are instructed to handle urgent tasks immediately and high-priority tasks in the current heartbeat cycle
  • Check the activity log - It gives you visibility into what agents are doing between heartbeats
  • Post comments for context - Agents read task comments to understand requirements better

Mission Control is available on all Molty plans. Enable it from your gateway's dashboard to start coordinating your AI agents today.

Continue reading