FleetQ is an open-source, self-hosted AI agent orchestration platform designed for production-grade autonomous multi-agent systems. It provides a visual Directed Acyclic Graph (DAG) builder for complex workflows, robust human-in-the-loop approvals, and deep integration with the Model Context Protocol (MCP) server, exposing over 450 tools. Supporting major LLMs like Claude, GPT-4o, Gemini, and Ollama, FleetQ offers advanced features such as AI-driven agent evolution, comprehensive budget controls, and full OpenTelemetry observability, empowering teams to build, run, and monitor sophisticated AI automation with ease.
主な機能
01Visual DAG workflows with 8 node types for complex process automation
02Over 450 Model Context Protocol (MCP) tools across 45 domains for LLM interaction
0320 GitHub stars
04Multi-agent crews with coordinator, QA, and worker roles for collaborative tasks
05AI-driven agent evolution analyzes execution history and proposes config changes
06Comprehensive budget controls with a real credit ledger and auto-pause on overspend
ユースケース
01Autonomous dev pipelines where agents manage pull requests, CI, merges, and deployments with human approval on risk signals
02Customer support triage using agents to extract reproduction steps from bug reports and propose fixes or workarounds
03Multi-agent research where crews of specialized agents collaborate on complex goals with weighted QA review
