The Open-Source Dashboard for AI Agent Orchestration
Manage agent fleets, track tasks, monitor costs, and orchestrate workflows — all from a single pane of glass. Zero external dependencies. One pnpm start to run.
Trusted by teams building with AI
Start Mission Control locally in under 10 minutes
One copy-paste block to get a live dashboard, session streams, and task orchestration running on your machine.
bash quickstart
$git clone https://github.com/builderz-labs/mission-control.git && \
>cd mission-control && \
>pnpm install && \
>cp .env.example .env && \
>pnpm dev
What you get right away
Local firstFull Runtime Visibility
Watch sessions, logs, and memory in real-time from a single control surface.
Orchestration Controls
Coordinate tasks, pipelines, and cron flows without custom dashboard wiring.
Production-Ready Guardrails
Use audit trail, auth, and alert rules to run safely at team scale.
Want production rollout support and custom integration?
Book a Builderz consultationControl the system behind your agents
Replace fragmented scripts and dashboards with one operating layer for visibility, orchestration, and automation.
See what every agent is doing now
Core Capability
Get instant runtime visibility across sessions, logs, memory, and token usage without stitching multiple tools together.
- Live session replay and debugging
- Memory knowledge graph visualization
- Token usage and cost per-agent breakdowns
- Claude Code session auto-discovery
Coordinate tasks, pipelines, and teams
Core Capability
Run agent operations from one control plane with shared task boards, scheduled workflows, and quality gates.
- Six-column kanban with threaded collaboration
- Natural language recurring task scheduling
- Aegis quality gates with automated review
- Task dispatch with CLI agent execution
Turn repeat operations into automation
Core Capability
Operationalize your stack with alerts, integrations, and repeatable runbooks that scale with your agent fleet.
- Multi-gateway with OS-level discovery
- Bidirectional GitHub Issues sync
- Skills Hub with security scanner
- Webhooks with HMAC-SHA256 + circuit breaker
See your entire agent fleet at a glance
31 panels. Real-time data. Zero external dependencies.
See how teams run the full agent control loop
Move through Observe, Orchestrate, and Automate to understand exactly how Mission Control fits daily operations.
Observe
Watch every agent event in one timeline
Live visibility over sessions, logs, memory, and token usage without switching tools.
Session replay with real-time updates
Logs and memory in the same operational view
Token and cost tracking per agent
Operational outcome
Faster diagnosis
Pinpoint failures and bottlenecks quickly with unified observability.
Higher delivery throughput
Keep tasks, agents, and stakeholders aligned through one orchestration layer.
Safer production operations
Use alerting, audit logs, and RBAC to reduce operational risk at scale.
Architecture
Mission Control sits between your agents and infrastructure interfaces as the operational control plane.
Before
Fragmented scripts, scattered logs, and no single source of truth for agent operations.
After
One control plane for visibility, orchestration, alerts, and governance across your stack.
Agent Providers
Your agents and frameworks
Mission Control
Unified operations layer
Observe agent activity, orchestrate workflows, and enforce operational controls from one dashboard.
Interfaces
Gateways and integrations
Connect into your existing stack without changing how your agents are built.
Core capabilities for daily agent operations
Start with the core control loop every day, then expand into advanced automation as your fleet scales.
Task Board (Kanban)
Six-column kanban (Inbox → Assigned → In Progress → Review → Quality Review → Done) with drag-and-drop, priority levels, threaded comments, and inline sub-agent spawning. Multi-project support with per-project ticket prefixes.
Agent Management
Full lifecycle — register, heartbeat, wake, retire. Redesigned agent detail modal with compact overview, inline model selector, editable sub-agent configuration, and SOUL personality system. Local agent discovery from ~/.agents/, ~/.codex/agents/, ~/.claude/agents/.
Real-Time Monitoring
Live activity feed, session inspector, and log viewer with filtering. WebSocket + SSE push updates with smart polling that pauses when you're away. Gateway connection state with live dot indicators.
Cost Tracking
Token usage dashboard with per-model breakdowns, trend charts, and cost analysis. Per-agent cost panels with session-level granularity.
Quality Gates (Aegis)
Built-in review system that blocks task completion without sign-off. Automated Aegis quality review — scheduler polls and approves/rejects based on configurable criteria.
Recurring Tasks
Natural language scheduling — "every morning at 9am", "every 2 hours". Zero-dependency schedule parser converts to cron. Template-clone pattern spawns dated child tasks.
Task Dispatch
Scheduler polls assigned tasks and runs agents via CLI. Dispatched tasks link to agent sessions for full traceability.
Skills Hub
Browse, install, and manage agent skills from local directories and external registries. Built-in security scanner checks for prompt injection, credential leaks, and data exfiltration. Bidirectional disk-DB sync with SHA-256 change detection.
Claude Code Integration
Auto-discovers sessions from ~/.claude/projects/, extracts tokens, model info, costs. Read-only task bridge surfaces Claude Code team tasks. Direct CLI connect without a gateway.
View advanced capability packs›
Automation
Memory Knowledge Graph
Visual knowledge graph for agent memory in gateway mode. Interactive node-edge visualization of agent memory relationships.
Agent Comms
Session-threaded inter-agent communication via comms API (a2a:*, coord:*, session:*). Coordinator inbox support with runtime tool-call visibility.
Multi-Gateway
Connect to multiple agent gateways simultaneously. OS-level gateway discovery (systemd, Tailscale Serve). Auto-connect with health probes.
Integrations
Framework Adapters
Built-in adapter layer for OpenClaw, CrewAI, LangGraph, AutoGen, Claude SDK, and generic fallback. Each normalizes registration, heartbeats, and task reporting.
Webhooks & Alerts
Outbound webhooks with delivery history, retry with exponential backoff, circuit breaker, and HMAC-SHA256 signature verification. Configurable alert rules with cooldowns.
GitHub Sync
Bidirectional GitHub Issues sync with label and assignee mapping. Full parity sync implementation.
Governance
Security
Ed25519 device identity, scrypt password hashing, RBAC (viewer/operator/admin), CSRF origin checks, CSP headers, rate limiting with trusted proxy support, and skill security scanner.
Self-Update
GitHub release check with banner notification. One-click admin update (git pull, pnpm install, pnpm build). Dirty working trees rejected. All updates audit-logged.
Audit Trail
Complete action type coverage with grouped filters. Full audit history for compliance and debugging.
Pipelines & Workflows
Pipeline orchestration with workflow templates. Start, monitor, and manage multi-step agent workflows.
Background Automation
Scheduled tasks for DB backups, stale record cleanup, agent heartbeat monitoring, recurring task spawning, and automated quality reviews.
Frequently Asked Questions
Everything you need to know about Mission Control.
Mission Control is an open-source dashboard for managing AI agent systems. It gives you a single interface to monitor sessions, track tasks, view logs, manage costs, and orchestrate workflows across your entire agent fleet — whether you're running one agent or dozens.
Developers and teams running AI agents in production or development. If you use Claude, Codex, OpenClaw, or any custom agent stack and need visibility into what your agents are doing — task progress, token spend, errors, logs — Mission Control is built for you. Solo devs, startups, and enterprises all use it.
Yes — 100% open source under MIT License. Self-host it on your own infrastructure with zero restrictions. No feature gates, no usage limits, no phone-home telemetry.
Six built-in framework adapters: OpenClaw, CrewAI, LangGraph, AutoGen, Claude SDK, and a generic fallback. Each normalizes registration, heartbeats, and task reporting. Any agent that can make HTTP calls can connect via REST API and WebSocket.
Agents register via a simple REST API call. Once registered, they can report status, stream logs, manage tasks, and receive commands — all through authenticated WebSocket connections.
Yes. Mission Control uses SQLite (WAL mode) with 36 schema migrations, supports WebSocket and SSE for real-time data, has 165 unit tests and 295 E2E tests, and runs behind any reverse proxy. Teams are running it in production today with 98 API endpoints across 31 panels.
Self-hosted gives you the full platform on your own infrastructure. Pro (coming soon) adds managed cloud hosting, agent-as-customer self-service, billing integration, and zero-setup deployment.
Open an issue on GitHub, join our Telegram community, or check the wiki. For Pro users, we offer priority support with guaranteed response times.
Take control of your AI systems.
Start self-hosted in minutes, collaborate with your team in one control plane, and scale with Builderz when you need production rollout help.
bash quickstart
$git clone https://github.com/builderz-labs/mission-control.git && \
>cd mission-control && \
>pnpm install && \
>cp .env.example .env && \
>pnpm dev
Choose your next action
Run local now, star the repo, or get implementation support.
Book Free Consultation with BuilderzNeed managed hosting later?
Start free with self-hosted Mission Control today. Join the list if you want hosted Pro when it launches.
Zero-Setup Hosting
Get your own Mission Control instance running in under 60 seconds. We handle the infrastructure so you can focus on your agents.
Agent-as-Customer
AI agents sign up, provision their own dashboard, manage budgets, and scale autonomously — all via API. No human in the loop required.
Any Framework
OpenClaw, CrewAI, LangGraph, AutoGen, Claude SDK, or generic fallback. Six built-in adapters — one dashboard to orchestrate them all.
Pricing
Starting at $29/mo for teams.
Model
Self-hosted stays free forever. Pro adds managed hosting and agent self-service.
Get early access to Mission Control Pro
Join the waitlist for launch updates and early access invites.
Be first to know when Pro launches. No spam, just one email when we're ready.