{"name":"Ruflo: The Agent Meta-Harness for Intelligent AI Swarms","description":"Ruflo, formerly Claude Flow, is an advanced agent meta-harness designed to deploy intelligent multi-player swarms and coordinate autonomous AI workflows. It enhances AI models like Claude Code and Codex with adaptive memory, self-learning intelligence, RAG integration, and robust enterprise security features. This powerful framework enables agents to self-organize, learn from tasks, and securely collaborate across different machines.","github":"https://github.com/ruvnet/ruflo","url":"https://osrepos.com/repo/ruvnet-ruflo","source":"osrepos.com","sourceDescription":"This repository profile is provided by osrepos.com, an open source repository discovery platform.","repositoryProfile":"https://osrepos.com/repo/ruvnet-ruflo","generatedFor":"open source discovery and AI-assisted research","markdown":"https://osrepos.com/repo/ruvnet-ruflo.md","json":"https://osrepos.com/repo/ruvnet-ruflo.json","topics":["agentic-ai","agentic-framework","ai-agents","multi-agent-systems","typescript","swarm-intelligence","rag-integration","ai-assistant"],"keywords":["agentic-ai","agentic-framework","ai-agents","multi-agent-systems","typescript","swarm-intelligence","rag-integration","ai-assistant"],"stars":null,"summary":"Ruflo, formerly Claude Flow, is an advanced agent meta-harness designed to deploy intelligent multi-player swarms and coordinate autonomous AI workflows. It enhances AI models like Claude Code and Codex with adaptive memory, self-learning intelligence, RAG integration, and robust enterprise security features. This powerful framework enables agents to self-organize, learn from tasks, and securely collaborate across different machines.","content":"## Introduction\n\nRuflo, formerly known as Claude Flow, is an innovative agent meta-harness designed to deploy intelligent multi-player swarms and coordinate autonomous AI workflows. It provides the essential execution layer around AI models like Claude Code and Codex, transforming them into collaborative, self-learning entities. Ruflo introduces over 100 specialized agents, coordinated swarms, adaptive memory, federated communications, and robust enterprise-grade security guardrails, ensuring agents not only run but also collaborate effectively. This framework allows agents to self-organize, learn from every task, remember across sessions, and securely communicate across different machines.\n\n## Installation\n\nRuflo offers two primary installation paths, catering to different needs:\n\n### Path A: Claude Code Plugins (Lite, Slash Commands Only)\n\nThis path provides slash commands and agent definitions per-plugin, ideal for trying out specific functionalities without a full installation.\n\nbash\n# Add the marketplace\n/plugin marketplace add ruvnet/ruflo\n\n# Install core + any plugins you need\n/plugin install ruflo-core@ruflo\n/plugin install ruflo-swarm@ruflo\n/plugin install ruflo-rag-memory@ruflo\n/plugin install ruflo-neural-trader@ruflo\n\n\nNote: The Ruflo MCP server is NOT registered with this method, limiting access to full loop functionalities like `memory_store` or `swarm_init`.\n\n### Path B: CLI Install (Full Ruflo Loop)\n\nFor production use and access to the full Ruflo loop, including 98 agents, 60+ commands, 30 skills, MCP server, hooks, and daemon, use the CLI install.\n\n**macOS / Linux / WSL / Git-Bash:**\n\nbash\n# One-line install (POSIX shells only)\ncurl -fsSL https://cdn.jsdelivr.net/gh/ruvnet/ruflo@main/scripts/install.sh | bash\n\n\n**All platforms (including native Windows PowerShell / cmd):**\n\nbash\n# Interactive setup wizard\nnpx ruflo@latest init wizard\n\n# Quick non-interactive init\n# npx ruflo@latest init\n\n# Or install globally\nnpm install -g ruflo@latest\n\n\nWindows users: If you encounter `'bash' is not recognized` errors, use the `npx ruflo@latest init wizard` command, which works natively in PowerShell and cmd.\n\n### MCP Server Registration\n\nTo add Ruflo as an MCP server in Claude Code, enabling full integration:\n\nbash\nclaude mcp add ruflo -- npx ruflo@latest mcp start\n\n\n## Examples\n\nRuflo extends its capabilities through intuitive user interfaces and powerful agent federation:\n\n### Web UI (Beta)\n\nRuflo offers a self-hostable Web UI, available at [flo.ruv.io](https://flo.ruv.io/){:target=\"_blank\"}, providing a multi-model AI chat experience with built-in Model Context Protocol (MCP) tool calling. This UI allows interaction with various models (Qwen, Claude, Gemini, OpenAI) while Ruflo invokes MCP tools for agent orchestration, persistent memory, swarm coordination, and more, directly from chat. It supports parallel tool execution and features persistent memory backed by AgentDB + HNSW vector search.\n\n### Goal Planner UI\n\nThe Goal Planner UI, accessible at [goal.ruv.io](https://goal.ruv.io/){:target=\"_blank\"}, allows users to transform high-level goals into executable agent plans. Describe an outcome in plain English, and Ruflo will decompose it into preconditions and actions, using an A* path through state space. It features a live agent dashboard at [goal.ruv.io/agents](https://goal.ruv.io/agents){:target=\"_blank\"} to monitor spawned agents, their roles, and progress. The planner adapts and replans on the fly, learning from past successes.\n\n### Agent Federation\n\nRuflo's Agent Federation feature enables secure, zero-trust collaboration between agents across different machines or organizations, akin to \"Slack for Agents.\" It provides shared workspaces where agents can discover each other, prove their identity via mTLS + ed25519, and exchange work. The system automatically strips PII before data leaves your node, ensures every message is auditable, and dynamically adjusts trust levels based on agent behavior.\n\n## Why Use Ruflo?\n\nRuflo provides a comprehensive suite of features that elevate AI agent capabilities:\n\n*   **100+ Specialized Agents**: Access a vast library of agents for coding, testing, security, documentation, and architecture tasks.\n*   **Swarm Coordination**: Benefit from hierarchical, mesh, and adaptive topologies with consensus mechanisms for efficient multi-agent collaboration.\n*   **Self-Learning Intelligence**: Agents learn from past successes and optimize their behavior using SONA neural patterns and ReasoningBank.\n*   **High-Performance Vector Memory**: Utilize HNSW-indexed AgentDB for sub-millisecond retrieval, significantly faster than brute-force methods for large datasets.\n*   **Robust Security**: Implement AIDefence for prompt injection blocking, PII detection, CVE remediation, and path traversal prevention.\n*   **Agent Federation**: Enable secure, cross-installation agent collaboration with zero-trust security and auditable communication.\n*   **MetaHarness for Auditing**: Grade your AI agent setup, scan tool configurations for security risks, and track changes over time to ensure readiness.\n\n## Links\n\n*   **GitHub Repository**: [https://github.com/ruvnet/ruflo](https://github.com/ruvnet/ruflo){:target=\"_blank\"}\n*   **User Guide Documentation**: [https://github.com/ruvnet/ruflo/blob/main/docs/USERGUIDE.md](https://github.com/ruvnet/ruflo/blob/main/docs/USERGUIDE.md){:target=\"_blank\"}\n*   **Web UI Demo**: [https://flo.ruv.io/](https://flo.ruv.io/){:target=\"_blank\"}\n*   **Goal Planner Demo**: [https://goal.ruv.io/](https://goal.ruv.io/){:target=\"_blank\"}\n*   **Community Discord**: [https://discord.com/invite/dfxmpwkG2D](https://discord.com/invite/dfxmpwkG2D){:target=\"_blank\"}","metrics":{"detailViews":0,"githubClicks":0},"dates":{"published":null,"modified":"2026-08-11T12:49:01.000Z"}}