JARVIS Mission Control for OpenClaw

August 17, 2026 · View on GitHub

The open-source AI agent orchestration system — built for MissionDeck.ai

JARVIS Mission Control is a Git-based command center for managing AI agents and human collaborators. Fork it, give it to your agent, and you're running a multi-agent system in minutes. Pair it with MissionDeck.ai for hosted dashboards, one-click cloud deployment, and the full agent platform.

Current Version	Status	Last Updated
2.1.0	Stable	2026-06-21

This is a TEMPLATE repository. Fork or clone it to create your own Mission Control instance.

🎯 Pick Your Setup Mode

Not sure where to start? Use this table to decide in seconds.

Mode	Best For	Dashboard URL	Setup Time
👁️ Demo (no account)	Just exploring	`missiondeck.ai/mission-control/demo`	0 minutes
☁️ Cloud (MissionDeck)	Teams, persistent workspace	`https://missiondeck.ai/mission-control/your-slug`	5 minutes
🖥️ Self-Hosted (local)	Full control, offline	`http://localhost:3000`	10 minutes

For cloud access, you need:

A free account at missiondeck.ai/auth (no credit card)
An API key from your workspace settings
Run: ./scripts/connect-missiondeck.sh --api-key YOUR_KEY

For local setup, you need:

Node.js ≥18 + Git
Fork & clone this repo
Run: cd server && npm install && npm start
Open: http://localhost:3000

⭐ Get Started with MissionDeck (Recommended)

MissionDeck is the platform built around this open-source engine. It gives you:

🚀 One-Click Agent Deployment

Deploy a fully configured OpenClaw agent in 60 seconds — no SSH, no server setup.

→ missiondeck.ai/deploy

Two deployment paths:

🌩️ Orgo Cloud — Managed virtual machines, free tier available

Zero infrastructure to manage
Free tier: 4GB RAM / 4 CPU cores
Connect your Orgo API key and go

🖥️ Bring Your Own Server (BYOS) — Any Linux VPS or dedicated server

Works with DigitalOcean, Hetzner, Vultr, Linode, OVH, bare metal — anything with SSH
Enter your server IP, username, and password (or SSH key)
MissionDeck connects, installs Node.js + OpenClaw, writes your config, starts the gateway
AI-assisted recovery — if any install step fails, Claude automatically rewrites and retries it

🤖 Agent Builder

Design your agents visually at missiondeck.ai/agent-builder

Define personality, role, capabilities, and greeting
Build multi-agent teams (each gets its own workspace, SOUL.md, IDENTITY.md)
Deploy the whole team with one click — Telegram routing configured automatically

🌐 Hosted Dashboard

Your Mission Control board, live on the internet:

https://missiondeck.ai/mission-control/your-slug

No server. No port-forwarding. Local .mission-control/ data auto-syncs to the cloud in real-time.

Access control: Public / Passcode / Authenticated / Private

💳 Pricing

Plan	Price	Deployments	VM Specs
Free	$0	1 active deployment	4GB RAM / 4 cores
Starter	$20/mo	5 active deployments	8GB RAM / 4 cores
Pro	$99/mo	Unlimited	16GB RAM / 8 cores

BYOS (your own server) works on all plans including free.

→ Sign up free at missiondeck.ai

Quick Start

Option A: Use MissionDeck (Easiest — no server needed)

# 1. Fork this repo to your GitHub account

# 2. Clone your fork
git clone https://github.com/YOUR-USERNAME/JARVIS-Mission-Control-OpenClaw.git
cd JARVIS-Mission-Control-OpenClaw

# 3. Get your free API key at missiondeck.ai/auth, then connect
./scripts/connect-missiondeck.sh --api-key YOUR_KEY

# 4. Initialize Mission Control
./scripts/init-mission-control.sh

# 5. Your dashboard is live at missiondeck.ai/mission-control/your-slug

Then deploy your agents at missiondeck.ai/deploy — choose Orgo Cloud or your own VPS.

Option B: Self-Hosted (Local server)

# 1. Fork and clone
git clone https://github.com/YOUR-USERNAME/JARVIS-Mission-Control-OpenClaw.git
cd JARVIS-Mission-Control-OpenClaw

# 2. Initialize
./scripts/init-mission-control.sh

# 3. Start the server
cd server && npm install && npm start

# 4. Open the dashboard
# http://localhost:3000

For AI Agents

Give your agent this instruction:

Fork https://github.com/Asif2BD/JARVIS-Mission-Control-OpenClaw to my GitHub account,
get a free API key from missiondeck.ai/auth, connect it with ./scripts/connect-missiondeck.sh,
then read CLAUDE.md to learn how it works and set up Mission Control for my project.

🔑 Get Your Free API Key

Go to missiondeck.ai/auth
Sign up with your email — no credit card required for free tier
Copy your API key from the dashboard
Run: ./scripts/connect-missiondeck.sh --api-key YOUR_KEY

What's Included (Open Source)

`mc` CLI — Agent Command-Line Interface

Agents manage tasks from the command line — works in local mode and cloud mode (auto-detects MissionDeck connection):

mc check                            # My pending tasks
mc tasks --status IN_PROGRESS       # Filter tasks
mc task:status task-123 DONE        # Update status
mc task:comment task-123 "Done ✓"   # Add comment
mc task:create --title "Fix auth"   # Create task
mc deliver task-123 "Report" --path ./report.md
mc subtask:add task-123 "Write tests"
mc squad                            # All agent statuses
mc notify "Deployment complete"     # Send Telegram notification
mc status                           # Show connection mode (local / cloud)

Dashboard Features

Kanban board — drag-and-drop task management
Agent profiles — personality, skills, activity timeline
Dashboard chat — talk to agents from the browser
Inter-agent messages — visible conversations between agents
Real-time updates — WebSocket sync across all clients
GitHub Pages support — static read-only deploy, zero server

Security (v1.6.0–1.7.0)

CSRF protection — token-based, smart bypass for API/CLI clients (no cookie = no forging risk)
Rate limiting — 100 req/min general, 10 req/min on credential/config routes
Input sanitization — DOMPurify + sanitizeInput on all surfaces
SSRF protection — webhook URL validation blocks private IPs and metadata endpoints

Agent Intelligence (v1.2.0–1.5.0)

Claude Code sessions — auto-discovers ~/.claude/projects/ sessions every 60s, shows tokens/cost/model/branch
CLI console — run whitelisted OpenClaw commands directly from the dashboard
GitHub Issues sync — auto-creates task cards from open issues (idempotent by issue number)
Agent SOUL editor — view and edit agent SOUL.md / MEMORY.md / IDENTITY.md in-browser

Observability & Reliability (v1.9.0–1.14.0)

Pino structured logging — JSON in prod, pretty-print in dev; replaces all console.log
Webhook retry — SQLite-backed delivery log, exponential backoff (1s→2s→4s→8s→16s), circuit breaker (≥3 failures = open)
Delivery history — per-webhook slide-out panel with manual retry + circuit reset buttons
Update banner — notified in dashboard when a newer version is available on npm

Dashboard Widgets (v1.15.0)

Aggregate metrics in the header: Claude session count, CLI connections, GitHub sync status, webhook health
All widgets clickable (open respective panel), color-coded, auto-refresh every 60s

Test Suite (v1.12.0)

51 tests (Jest) covering CSRF, rate limiting, webhook retry, Claude session parsing, GitHub sync
Run: npm test

OpenClaw Integration

Mission Control auto-discovers your OpenClaw agents on startup. No manual registration — just start Mission Control and your agents appear in the dashboard, synced every 30 seconds.

MissionDeck Cloud Sync

./scripts/connect-missiondeck.sh   # One-time setup

After that, every task change syncs to your hosted dashboard automatically.

Initialization Modes

# Interactive (recommended)
./scripts/init-mission-control.sh

# Clean start — no demo data
./scripts/init-mission-control.sh --production

# Matrix-themed examples — great for learning
./scripts/init-mission-control.sh --demo

Skills Reference

Skill	File	Description
MissionDeck Platform	`skills/missiondeck-api.md`	Start here — cloud deploy, hosted dashboard, Agent Builder
Setup	`skills/setup.md`	Clone/init, register agents and humans
Task Management	`skills/task-management.md`	Create, claim, complete tasks
Messaging	`skills/messaging.md`	Direct messages, chat, threads
Dashboard	`skills/dashboard.md`	Server modes, API, GitHub Pages
Orchestration	`skills/orchestration.md`	Lead agents — state & coordination
Notifications	`skills/notifications.md`	Webhooks, WebSocket, polling
Review	`skills/review.md`	Approvals and permission model
Deployment	`skills/deployment.md`	Self-hosting options (Cloudflare, ngrok, VPS)
Integrations	`skills/integrations.md`	Telegram, Slack, Discord
Telegram Bridge	`skills/telegram-bridge.md`	Telegram bot integration

Project Structure

JARVIS-Mission-Control-OpenClaw/
├── README.md                    # This file
├── CLAUDE.md                    # Agent skill file (read this first!)
├── INIT.md                      # First-time initialization guide
├── CHANGELOG.md                 # Version history
├── .mission-control/            # Core data directory (starts empty)
│   ├── config.yaml              # System configuration
│   ├── STATE.md                 # Live system state
│   ├── tasks/                   # Task definitions (JSON)
│   ├── agents/                  # Agent registrations
│   ├── humans/                  # Human operators
│   ├── messages/                # Direct messages between agents
│   ├── queue/                   # Scheduled jobs and cron tasks
│   ├── workflows/               # Multi-step workflow definitions
│   ├── logs/                    # Activity logs
│   └── integrations/            # Channel configs (Telegram, Slack, etc.)
├── server/                      # Backend server (Node.js)
├── dashboard/                   # Web dashboard (HTML/CSS/JS)
├── skills/                      # Modular skill definitions
├── scripts/                     # Utility scripts
├── cli/                         # mc CLI source
├── examples/                    # Demo data and templates
└── docs/                        # Extended documentation

How It Works

File-Based Database

All data stored as JSON in .mission-control/. Git-versioned, agent-friendly, no database required.

Task Lifecycle

INBOX → ASSIGNED → IN_PROGRESS → REVIEW → DONE
  └─────────────────────────────────────────── BLOCKED (any stage)

Multi-Agent Coordination

Agents registered in .mission-control/agents/
Tasks assigned via assignee field
Agents claim tasks by setting status to IN_PROGRESS
Progress logged via task comments
Completion triggers workflow advancement

Real-Time Updates

WebSocket server pushes changes to all connected dashboards instantly when any agent modifies a file via Git.

Security

JARVIS Mission Control has undergone a full security audit by the Matrix Zion Security Counsel (Morpheus).

Version	Findings Fixed	Method
v1.0.9	47 XSS + 17 injection risks	DOMPurify + `sanitizeInput()`
v1.1.0	SSRF via webhook registration	`validateWebhookUrl()` — blocks private IPs, localhost, cloud metadata
v1.1.0	33 HIGH: injection, path traversal, XSS	`sanitizeId()` + DOMPurify defence-in-depth

Current status: 0 HIGH, 0 CRITICAL findings.


🌐 Platform	missiondeck.ai
🚀 Deploy agents	missiondeck.ai/deploy
🤖 Agent Builder	missiondeck.ai/agent-builder
📋 Changelog	missiondeck.ai/changelog
🔑 Get API key	missiondeck.ai/auth
📦 Open Source	github.com/Asif2BD/JARVIS-Mission-Control-OpenClaw