ruvnet/ruflo

⭐ 48,191  ·  TypeScript  ·  GitHub Repo

🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade architecture, self-learning swarm intelligence, RAG integration, and native Claude Code / Codex Integration

agentic-ai agentic-framework agentic-rag agentic-workflow agents ai-agent ai-assistant ai-coding

1-Sentence Summary

The leading agent orchestration platform for Claude, enabling self-organizing AI swarms with zero-trust federation and self-learning memory.

🔥 Key Capabilities & USP

• Multi-Agent Swarm Orchestration — Coordinate 100+ specialized AI agents that self-organize into teams and autonomously route tasks. Solves the pain point of managing complex, multi-step workflows that exceed a single agent's capacity.
• Self-Learning & Memory — Agents learn from every task, remember across sessions, and optimize behavior using vector databases and graph-based retrieval. Eliminates the need to re-train or re-contextualize agents for recurring work.
• Zero-Trust Federation — Agents on different machines and organizations securely discover, authenticate, and communicate without data leakage. Solves the critical pain point of cross-boundary agent collaboration in enterprise environments.
• 32-Plugin Ecosystem — Modular plugins for code quality, security, DevOps, architecture, and domain-specific use cases (trading, IoT) with Wasm extensibility. Eliminates the need to build custom integrations from scratch.
• Native Claude Code / Codex Integration — Works as a plugin within Claude Code or as a full CLI, providing a complete "nervous system" for Claude. USP: No other platform offers this depth of native integration combined with self-learning swarms and enterprise federation.

Technical Architecture

Component	Technology / Role
Core Engine	Rust-based AI engine (Cognitum.One) for embeddings, memory, and plugins
Orchestration	Router → Swarm → Agents loop with self-learning feedback
CLI	npx ruflo@latest for full platform access
MCP Server	Model Context Protocol server for Claude integration
Vector Database	RuVector for persistent agent memory
Hooks System	Event-driven hooks for workflow automation
Daemon	Background process for continuous agent operation
LLM Providers	Multi-provider support (Anthropic Claude primary)
Federation	Zero-trust protocol for cross-machine agent communication

Quick Start Guide

Plugin install (lite, within Claude Code):

/plugin marketplace add ruvnet/ruflo
/plugin install ruflo-core@ruflo
/plugin install ruflo-swarm@ruflo
/plugin install ruflo-autopilot@ruflo
/plugin install ruflo-federation@ruflo

CLI install (full, production-grade):

curl -fsSL https://cdn.jsdelivr.net/gh/ruvnet/ruflo@main/scripts/install.sh | bash
npx ruflo@latest init wizard
npm install -g ruflo@latest

MCP Server setup:

claude mcp add ruflo -- npx ruflo@latest mcp start

Pros, Cons & Use Cases

Pros

• Zero workspace pollution — Plugin path adds no files to your project
• Production-ready — Full CLI with daemon, MCP server, and hooks system
• Extensive plugin ecosystem — 32 pre-built plugins covering code quality, security, DevOps, and domain-specific use cases
• Self-learning architecture — Agents improve over time without manual retraining
• Enterprise security — Zero-trust federation for cross-organization collaboration

Cons

• Two-tier functionality — Plugin install path lacks MCP server and full capabilities compared to CLI install
• Claude dependency — Requires Claude Code as the underlying platform; not LLM-agnostic
• Learning curve — Complex ecosystem with multiple components (CLI, MCP, daemon, plugins) may overwhelm new users

Who should NOT use this?

• Teams not using Claude Code or Anthropic's ecosystem
• Simple single-agent use cases that don't require orchestration
• Projects that cannot accommodate the complexity of a multi-agent framework
• Organizations with strict policies against external AI agent federation

Ideal Use Cases

• Enterprise CI/CD pipelines — Deploy swarms of specialized agents for code review, security scanning, and deployment automation
• Cross-team collaboration — Federated agents across different departments or organizations working on shared codebases
• Complex research workflows — Multi-agent systems that need to learn and adapt from ongoing experiments
• Autonomous DevOps — Self-organizing agent teams that manage infrastructure, monitoring, and incident response
• AI-powered development assistants — Custom agent swarms for specific domains (trading, IoT, architecture)

Community & Activity

With 48,191 stars and active development through May 2026, Ruflo has clearly struck a nerve in the AI engineering community. This is not a side project — it's a rapidly growing platform with serious momentum. The combination of Claude Code native integration, self-learning swarms, and enterprise federation has resonated strongly with developers building production AI systems. The active plugin ecosystem and regular updates signal a project that's being actively maintained and evolved, making it a safe bet for teams looking to invest in a long-term agent orchestration solution.