Mission Control
A self-hosted control plane for running several agents: task dispatch, run review, spend tracking, connected via MCP, CLI or REST
High risk
We rate an entry high when the tool writes to external systems, handles money, production databases or secrets, or runs arbitrary commands. The CLI installs it only with your consent.
Why this level
- Stores API keys and credentials for controlling several agents
- Lets callers register agents and dispatch tasks to them programmatically
- The README explicitly warns to keep the instance on a trusted network without a TLS reverse proxy
Install
Manual install
git clone https://github.com/builderz-labs/mission-control.git
cd mission-control
bash install.sh --localLocal install from source, needs Node.js 22+ and pnpm.
This is third-party code. Review the repository files before installing.
What it does
Mission Control is a separate self-hosted app with a web dashboard and a SQLite database that sits above agent runtimes like OpenClaw, Claude Code or Codex without replacing their reasoning loop. It registers agents, assigns and dispatches tasks, tracks runs through a quality review gate, keeps memory and a skills registry, and tallies spend and token use. An agent connects over the REST API with an API key, through the built-in CLI, or through an MCP server added to Claude Code with one command. It fits when several agents or runtimes work in parallel and it is hard to tell who owns what and what already passed review.
Who it is for. For anyone running several agents or runtimes in parallel who wants one dashboard for tasks, review and spend.
Good fit when
- Several agents work in parallel and it is unclear who owns what and what already passed review
- You want unified spend and token tracking across all agents and runtimes
- You need to dispatch tasks to an agent programmatically over REST, CLI or MCP and watch the queue
Not a fit when
- A single agent on a single machine already stays understandable from its native CLI
- You need a managed multi-tenant SaaS rather than a self-hosted control plane
- The project is still alpha and you cannot tolerate schema or API changes between releases
Example request
Register me as agent scout in Mission Control and show me my task queueLimitations
The project is alpha, schema and API can change between releases. It runs as a separate Next.js service with SQLite, needs a first admin account and an API key to store. Live session messaging needs a connected runtime gateway, some features are unavailable without one.
How to disable. Stop the Mission Control container or process and remove the MCP connection with claude mcp remove mission-control.
MCP
- Transport
- stdio
- Authentication
- API key
| Environment variables | |
|---|---|
| MC_URL required | Address of the deployed Mission Control instance. |
| MC_API_KEY required, secret | API key from Settings after the first setup. |
Security check
- Stores API keys and credentials for controlling several agents
- Lets callers register agents and dispatch tasks to them programmatically
- The README explicitly warns to keep the instance on a trusted network without a TLS reverse proxy
README in short
The README frames Mission Control as a self-hosted control plane for operators, sitting above OpenClaw, Claude Code, Codex and other runtimes. It lists its scope: tasks with a quality review gate, agent registration and presence, an activity stream, memory and a skills registry, roles and audit, plus Web UI, CLI, MCP server and REST API interfaces. It explicitly states when the tool is the wrong fit: a single agent on one machine, a need for managed SaaS, or an inability to tolerate alpha status. Install is via docker compose or a source build with Node.js and pnpm, MIT licensed.
FAQ
Does Mission Control replace the agent's reasoning loop?
No, the README states the control plane sits above runtimes and does not replace their reasoning and tool loop, it gives operators one place to see and govern the work.
Is a runtime gateway required?
Not for tasks, spend tracking or skills, but live session messaging needs a connected gateway for that specific runtime.
Related
An MCP server with n8n node and template knowledge: the agent picks nodes, validates configs and, with API access, creates workflows in your n8n
Zapier's official MCP plugin: the agent gets actions across thousands of apps through your Zapier account
Awesome Claude Skills by Composio
Awesome Claude Skills
A curated list of Claude skills and plugins, plus Composio's own automation skills for 78 SaaS apps
A plugin and CLI catalog for agents: generates command-line interfaces for GUI apps like GIMP and Blender and installs ready ones via CLI-Hub