Mission Control

Mission Control is an open-source AI agent orchestration dashboard for managing agent fleets, tracking tasks, monitoring costs, and coordinating workflows from a single control surface. It is described as self-hosted in minutes and uses a local-first setup with zero external dependencies.

Its interface centers on three main functions: observe, orchestrate, and automate. For observability, it shows sessions, logs, memory, and token usage in real time, with live session replay, debugging, a memory knowledge graph visualization, and per-agent cost breakdowns. For orchestration, it provides a six-column kanban board with threaded collaboration, drag-and-drop task handling, recurring task scheduling in natural language, and quality gates that block completion without review. It also supports agent lifecycle management, including register, heartbeat, wake, and retire, plus task dispatch, pipelines, and cron flows.

The dashboard includes 31 panels and 98 API endpoints, and it offers interfaces such as WebSocket, REST API, SSE, and webhooks. The page also mentions bidirectional GitHub Issues sync, a skills hub with a security scanner, and a webhook system with HMAC-SHA256 and a circuit breaker. It supports framework adapters for OpenClaw, CrewAI, LangGraph, AutoGen, and Claude SDK, and it includes local agent discovery from ~/.agents/, ~/.codex/agents/, and ~/.claude/agents/. The runtime views include a live activity feed, session inspector, log viewer with filtering, and gateway connection state indicators.

Mission Control is aimed at startups, product teams, automation agencies, and internal AI platforms. It is MIT licensed, self-hosted, and explicitly states that there is no forced telemetry and an open API. The page also references GitHub stars and forks, and invites users to book a Builderz consultation for production rollout support and custom integration.