mcp server

robofang

openclaw / openfang inspired experiment. adds mcp server fleet and robotics

sandraschi/robofangcommunitylistedunreviewed

Use tool

Set up

Runs on your computer. Pick your agent and operating system.

Before you start

uv (`curl -LsSf https://astral.sh/uv/install.sh | sh`)

~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "robofang": {
      "command": "uvx",
      "args": [
        "robofang"
      ]
    }
  }
}

same format Cursor, GitHub Copilot CLI, Junie, JetBrains AI Assistant, Augment (Auggie CLI), Factory Droid, Kiro, Devin CLI +12

Values in angle brackets are placeholders. Restart Claude Desktop after editing its config.Claude Desktop docs

README

synced Oct 21, 2026

🪝 RoboFang v1.8.0 (Beta)

📖 Installation Guide — quick start, manual setup, and troubleshooting

The Command Center for AI Agent Fleets and Real-World Robotics.

[!WARNING] RoboFang is currently in Beta Status. Version v1.8.0 introduces significant architectural changes for physical integration and security. Expect rapid iteration as we move towards a stable release.

RoboFang is a secure hub that connects your AI agents (MCP servers) to the physical world. It provides a local-first environment where agents can safely control smart home devices, manage professional software tools, and operate robotic hardware. It prioritizes privacy, local execution, and safety.

Quick Start

git clone https://github.com/sandraschi/robofang
cd robofang
just

This opens an interactive dashboard showing all available commands. Run just bootstrap to install dependencies, then just serve or just dev to start.

Ports

Port	Service
10870	Hub web dashboard (Vite React) + bridge API
10871	Bridge server (FastAPI + MCP streamable HTTP)
10872	Supervisor (heartbeat, auto-restart, metrics)
10873	robofang-mcp thin server frontend

📖 Comprehensive Documentation

🛠️ Getting Started

Installation Guide: System requirements and setup instructions.
Quick Start: Your first mission with the RoboFang Council.
CLI Toolbelt: Managing the fleet from the command line.
Onboarding: Setting up communication channels and credentials.

🏗️ System Architecture

Agentic Mesh: High-level overview of the federated node architecture.
Cognitive Loop: Deep dive into the foreman/labor/audit reasoning engine.
OpenClaw++ Standards: The underlying protocols for autonomous coordination.
Philosophy: Why we build for sovereign agency.

🕸️ The Mesh (Software & Tools)

MCP Fleet: Discovering and managing model context protocol servers.
Skill Manager: Developing and deploying custom "hands" for specific tasks.
Federation: Bridging multiple RoboFang nodes across Tailscale.
MemOps: Long-term memory and knowledge graph integration.

🛡️ Security & Safety

Safety Logic: How the Mission Loop prevents non-deterministic failures.
Dark Twin (DTU): Using filesystem shadows for safe staging.
Security Integrations: DefenseClaw, OpenShell, Bastio — planned; honesty contract.
Competitive landscape: vs OpenClaw, OpenFang, OpenManus.
Chat UX: Chat-first hub (what users know).
Next priorities: Maintainer backlog P0–P3.

🤖 Physical Agency (Robotics)

Robotics Overview: Standardizing actuator mapping and motion control.
Robot Safety & Hazards: Crucial protocols for protecting your physical space.
Noetix Bumi Android: Guide to operating the flagship 1m humanoid substrate.
Yahboom Baseline: Entry-level support for Raspbot and mapping platforms.
Snark T-800: Decommissioned status and critical safety warnings.

🚀 Quick Start

# 1. Clone & Sync
git clone https://github.com/sandraschi/robofang
cd robofang
uv sync

# 2. Secure Launch
uv run python -m robofang.main

Engineering for sovereign agency. Every line of code is a step towards physical presence.