Yggdrasilbuilderz-labs/mission-control
↗ GitHubSelf-hosted AI agent orchestration platform: dispatch tasks, run multi-agent workflows, monitor spend, and govern operations from one mission control dashboard.
3,676
Stars
632
Forks
20
Watchers
15
Open Issues
Safety Rating A
The repository presents as a legitimate open-source AI agent orchestration platform with a well-structured codebase, comprehensive documentation, security hardening guidance, a disclosed vulnerability policy (SECURITY.md), and standard MIT licensing. No hardcoded secrets, malicious code patterns, suspicious dependencies, or prompt injection attempts were detected in the available content. The project explicitly advises users to change default credentials and deploy behind TLS, which reflects good security practice rather than a red flag.
ℹAI-assisted review, not a professional security audit.
AI Analysis
Mission Control is an open-source, self-hosted AI agent orchestration dashboard built with Next.js and TypeScript. It provides a single-pane-of-glass interface for managing AI agent fleets, dispatching tasks, tracking token costs, coordinating multi-agent workflows, and enforcing security policies. The platform runs on SQLite with no external dependencies (no Redis or Postgres required), offers 101 REST endpoints, real-time WebSocket/SSE updates, role-based access control, and integrates with frameworks including OpenClaw, CrewAI, LangGraph, AutoGen, and the Claude SDK.
Use Cases
- Managing and monitoring fleets of AI agents from a centralized dashboard
- Dispatching and tracking tasks across multiple agents using a Kanban-style board
- Running multi-agent workflows with quality gates and automated orchestration rules
- Tracking LLM token usage and costs per model and session
- Auditing agent security posture including secret detection, MCP call auditing, and trust scoring
- Scheduling recurring agent tasks using natural language cron expressions
- Integrating Claude Code sessions and team tasks into a unified observability surface
- Self-hosting an agent control plane with zero external infrastructure dependencies
Tags
Security Findings (4)
No hardcoded secrets detected in the repository content provided. The README explicitly states that credentials are auto-generated at install time, and the .env.example pattern is used for configuration. Default credential warnings are provided prominently in the security section.
No prompt injection attempts detected. The README content is straightforward documentation with no embedded instructions targeting AI analysts.
No malicious code patterns detected from the repository metadata and README content provided. The codebase is TypeScript/Next.js with standard dependencies (better-sqlite3, Zustand, Zod, Recharts, Vitest, Playwright).
No specific CVEs are identifiable from the metadata alone. The stack (Next.js 16, React 19, TypeScript 5.7, Zod 4, Zustand 5) uses current major versions. No obviously vulnerable pinned dependencies are visible from the README tech stack table.
Project Connections
clawguard
ClawGuard provides real-time activity monitoring and a security kill switch for the OpenClaw/Clawdbot gateway ecosystem, while Mission Control provides the broader orchestration dashboard. The two tools address complementary aspects of agent observability and security within overlapping gateway ecosystems.
CodexSkillManager
Mission Control's Skills Hub manages agent skills from registries including ClawdHub and ~/.codex/skills, which directly overlaps with the directories CodexSkillManager manages. The two tools could be used together — CodexSkillManager for local macOS skill management and Mission Control for fleet-wide skill deployment and security scanning.
xyops
Both xyOps and Mission Control are self-hosted orchestration/automation dashboards with job scheduling, monitoring, alerting, and webhook support. xyOps is general-purpose workflow automation, while Mission Control is specifically designed for AI agent orchestration.
agentfield
Both AgentField and Mission Control serve as backend control planes for AI agents, offering agent registration, task routing, async execution, and observability. AgentField is Go-based and API/microservice-oriented; Mission Control is TypeScript/Next.js and dashboard-oriented.
ClawWork
ClawWork benchmarks AI agents on real-world tasks using the OpenClaw gateway with a FastAPI+React dashboard, while Mission Control provides the production orchestration layer for those same agent types. ClawWork-evaluated agents could be deployed and monitored via Mission Control.