escapeboy/agent-fleet-o

> **Self-hosted mission control for AI agents.** Build, run, and monitor autonomous multi-agent systems with a visual DAG builder, human-in-the-loop approvals, MCP server integration, and full audit trail. Works with Claude, GPT-4o, Gemini, Ollama, Codex, Claude Code, and any OpenAI-compatible LLM.
**Keywords:** AI agents · agent orchestration · MCP server · Model Context Protocol · LangGraph alternative · CrewAI alternative · n8n for AI · Claude agents · LLM workflow · autonomous agents · agent framework · AI automation · self-hosted
☁️ **Prefer managed?** Try **FleetQ Cloud** — zero setup, free tier.
⭐ **Like the project?** Give it a star on GitHub — it helps others find FleetQ.
---
Table of Contents
- Why FleetQ?
- Key Concepts
- Screenshots
- Features
- Use Cases
- How FleetQ compares
- Quick Start
- Authentication
- Configuration
- SSH Host Access
- Architecture
- MCP Server (670+ tools)
- Tech Stack
- Contributing
- Changelog
---
Why FleetQ?
Most agent frameworks give you a Python notebook. FleetQ gives you a **production platform**.
- 🧩 **670+ MCP tools across 60+ domains** — every feature is exposed via Model Context Protocol, so any LLM (Claude Desktop, Cursor, ChatGPT, local agents) can drive the platform programmatically. New in 1.27: web UIs for previously headless capabilities (agent sessions, release signing keys, drift & eval monitors, broadcasts, test suites, CSV import); **eight outbound chat channels** as first-class drivers; the **Agentic AI Flywheel** (self-growing eval set + drift/production monitors); **policy-governed autonomy** (versioned per-agent policies + replay); **cost-aware orchestration** and **Return on Cognitive Spend (ROCS)** metrics.
- 🔁 **Visual DAG workflows** with 8 node types (agent, conditional, human-task, switch, dynamic-fork, do-while, compensation, sub-workflow) — no Python glue code.
- 👥 **Multi-agent crews** with coordinator/worker/reviewer roles, weighted QA scoring, and cross-validation.
- 🛡️ **Real-World Action governance** — assistant tool calls, integration writes, and git pushes route through a per-tier risk policy (auto / ask / reject for low / medium / high). Approvals auto-execute. Audit trail attached.
- 💰 **Budget controls** with a real credit ledger, pessimistic locking, and auto-pause on overspend — not just token counters.
- 🧠 **Agent evolution** — LLM analyzes execution history and proposes config changes you approve with one click.
- ⚙️ **BYOK + Local LLMs** — Anthropic, OpenAI, Google, plus Ollama, LM Studio, vLLM, Codex, Claude Code. Zero vendor lock-in.
- 🔒 **Production-grade** — tenant isolation, encrypted credential vault, HMAC webhooks, SSRF guards, circuit breakers, audit trail.
- 📊 **OpenTelemetry observability** — structured error codes (gRPC-canonical), deadline propagation, distributed tracing. Jaeger UI one-command away. Per-team OTLP collector endpoints for BYO observability.
- 📈 **Live team graph** — Cytoscape.js force-directed visualization of agents, humans, and crews. Real-time updates via Laravel Reverb WebSockets.
- 🏠 **Self-host or cloud** — MIT-friendly AGPLv3 license, runs on Docker Compose, or use FleetQ Cloud.
Key Concepts
| Concept | What it is | When to use |
|---|---|---|
| **Agent** | A configured AI personality with role, goal, backstory, skills, and tool access | The basic unit — one agent per specialized task |
| **Skill** | A reusable LLM prompt, rule, connector, or GPU compute call | When multiple agents need the same capability |
| **Experiment** | A stateful run through a 20-stage pipeline (scoring → planning → building → executing → evaluating) | Any non-trivial agent task with lifecycle |
| **Crew** | A team of agents working on one goal (sequential, parallel, hierarchical, adversarial, fanout, chat-room) | Multi-perspective tasks or when you need review/QA |
| **Workflow** | A visual DAG template (reusable across experiments) with branching, loops, human-tasks | Recurring processes — CI/CD, content pipelines, QA flows |
| **Project** | A continuous (cron-scheduled) or one-shot container for experiments, with budget + milestones | Long-running initiatives, scheduled agent work |
| **Signal** | An inbound event (webhook, RSS, email, bug report, GitHub issue) that can trigger agents | Event-driven automation |
| **MCP Tool** | A programmatic action any LLM can call to query or mutate the platform | Expose FleetQ to external agents (Claude, Cursor, etc.) |
Screenshots
**Dashboard**
KPI overview with active experiments, success rate, budget spend, and pending approvals.