open-computer-use
by LLmHub-dev
The Open Framework for autonomous virtual computer agents at scale, fully open-source, safe, auditable, and production-ready.
316stars
37forks
TypeScript
Added 12/27/2025
AI Agentsagentagentic-frameworkaiai-agentsanthropicclaudeclaude-codecli-toolscomputer-controlcomputer-usecomputer-use-agentfull-stackgemini-aigemini-apigoogle-apigui-automationllmllm-agentsopenaisaas-application
Installation
# Add to your Claude Code skills
git clone https://github.com/LLmHub-dev/open-computer-useGetting Started
Guides for using ai agents skills like open-computer-use.
- Caveman: Cut Claude Token Use by 65%How agent-side prompt compression works, when to use it, and when not to.
- What is an AI Skills Marketplace?Definitions, how marketplaces work, and how to choose between them in 2026.
- Getting Started with AI SkillsFirst-time install walkthrough for Claude Code, Codex CLI, and ChatGPT.
Security ReportVerified
Last scanned: 5/30/2026
{
"issues": [
{
"type": "npm-audit",
"message": "@ai-sdk/anthropic: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/azure: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/google: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/mistral: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/openai: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/openai-compatible: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/perplexity: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/provider-utils: @ai-sdk/provider-utils has an Uncontrolled Resource Consumption issue",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/react: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/ui-utils: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@ai-sdk/xai: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "@azure/identity: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "@azure/msal-node: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "@openrouter/ai-sdk-provider: Vulnerability found",
"severity": "low"
},
{
"type": "npm-audit",
"message": "ai: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "file-type: file-type affected by infinite loop in ASF parser on malformed input with zero-size sub-header",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "jsondiffpatch: jsondiffpatch is vulnerable to Cross-site Scripting (XSS) via HtmlFormatter::nodeBegin",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "next: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "next-intl: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "postcss: PostCSS has XSS via Unescaped </style> in its CSS Stringify Output",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "qs: qs has a remotely triggerable DoS: qs.stringify crashes with TypeError on null/undefined entries in comma-format arrays when encodeValuesOnly is set",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "uuid: uuid: Missing buffer bounds check in v3/v5/v6 when buf is provided",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "ws: ws: Uninitialized memory disclosure",
"severity": "medium"
}
],
"status": "PASSED",
"scannedAt": "2026-05-30T14:59:49.470Z",
"npmAuditRan": true,
"pipAuditRan": true
}💻 Open Computer Use - Autonomous Computer Using Agents at Scale
Your AI Agent That Actually Uses Computers Like Humans Do
Open Computer Use is an open-source platform that gives AI agents real computer control through browser automation, terminal access, and desktop interaction. Built for developers who want to create truly autonomous AI workflows.
Preview
✨ What Makes This Special?
Unlike traditional AI assistants that only talk about tasks, Open Computer Use enables AI agents to actually perform them by:
- 🌐 Browsing the web like a human (search, click, fill forms, extract data)
- 💻 Running terminal commands and managing files
- 🖱️ Controlling desktop applications with full UI automation
- 🤖 Multi-agent orchestration that breaks down complex tasks
- 🔄 Streaming execution with real-time feedback
- 🎯 100% open-source and self-hostable
"Computer use" capabilities similar to Anthropic's Claude Computer Use, but fully open-source and extensible.
🎬 See It In Action
Browser Automation
AI agent searching, navigating, and interacting with websites autonomously
▶️ Watch: AI Agent Browsing and Playing
Terminal Operations & Development
Executing commands, managing files, and running complex workflows
▶️ Watch: Quant Trading & Research on QuantConnect
Multi-Agent Orchestration
Complex tasks broken down and executed by specialized agents
▶️ Watch: Building Nvidia Options Dashboard
Advanced Features
Human-in-the-loop control and intelligent collaboration
▶️ Watch: AI Agent with Human Intervention
🎯 Core Capabilities
🌐 Browser Agent
- Search-first strategy using Google Search API
- Smart web navigation with automatic form filling
- Element detection and intelligent clicking
- Multi-tab management for parallel workflows
- Page context extraction for AI understanding
- Screenshot capture for visual verification
💻 Terminal Agent
- Command execution in isolated environments
- File operations (read, write, edit, delete)
- Directory management with full control
- Script execution (Python, Node.js, bash)
- Package installation and environment setup
- Output streaming with real-time feedback
🖱️ Desktop Agent
- UI element detection using computer vision
- Mouse and keyboard control for any application
- Window management (focus, resize, arrange)
- Screenshot analysis for context awareness
- OCR capabilities for text extraction
- Cross-platform support (Linux desktop)
🤖 Multi-Agent System
- Task decomposition by AI planner
- Sequential execution with context passing
- Specialized agents for different capabilities
- Error handling with automatic retries
- User interaction when clarification needed
- Execution reports with detailed summaries
🏗️ Architecture
┌─────────────────────────────────────────────────────────────────┐
│ Frontend (Next.js 15) │
│ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │
│ │ Chat UI │ │ Model │ │ VM │ │
│ │ Components │ │ Selection │ │ Management │ │
│ └──────────────┘ └──────────────┘ └──────────────┘ │
└─────────────────────────────────────────────────────────────────┘
▼
┌─────────────────────────────────────────────────────────────────┐
│ Backend API (FastAPI) │
│ ┌──────────────────────────────────────────────────────────┐ │
│ │ Multi-Agent Executor Service │ │
│ │ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │ │
│ │ │ Planner │→ │ Browser │→ │ Terminal │ │ │
│ │ │ Agent │ │ Agent │ │ Agent │ │ │
│ │ └─────────────┘ └─────────────┘ └─────────────┘ │ │
│ └──────────────────────────────────────────────────────────┘ │
│ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │
│ │ WebSocket │ │ Database │ │ Billing │ │
│ │ VM Control │ │ Service │ │ Service │ │
│ └──────────────┘ └──────────────┘ └──────────────┘ │
└─────────────────────────────────────────────────────────────────┘
▼
┌─────────────────────────────────────────────────────────────────┐
│ Docker VM (Ubuntu 22.04 + XFCE) │
│ ┌──────────────────────────────────────────────────────────┐ │
│ │ Chrome Browser │ Terminal │ Desktop Apps │ Tools │ │
│ └──────────────────────────────────────────────────────────┘ │
│ ┌──────────────────────────────────────────────────────────┐ │
│ │ WebSocket Agent Server (Port 8080) │ │
│ │ VNC Server (Port 5900) │ │
│ └──────────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘
🚀 Quick Start
Prerequisites
- Node.js 20+ and npm
- Python 3.10+ and pip
- Docker and Docker Compose
- Supabase account (free tier works)
- API keys for AI providers (OpenAI, Anthropic, etc.)
1. Clone the Repository
git clone https://github.com/LLmHub-dev/open-computer-use.git
cd open-computer-use
2. Set Up Supabase Database
Create Supabase Project
- Go to Supabase and create a new project
- Wait for the project to finish setting up
- Go to Project Settings → API to get your keys
Run Database Schema
Execute the schema to create all required tables:
# Option A: Using Supabase Dashboard
# 1. Go to SQL Editor in your Supabase dashboard
# 2. Copy contents of supabase/schema.sql
# 3. Paste and run the SQL
# Option B: Using Supabase CLI (recommended)
npm install -g supabase
supabase login
supabase link --project-ref your-project-ref
supabase db push
Or manually run the schema file:
psql -h db.your-project.supabase.co -U postgres -d postgres -f supabase/schema.sql
This creates all necessary tables:
- 👤 Users & Auth: users, user_preferences, user_keys
- 💬 Chat System: chats, messages, chat_participants, chat_attachments
- 🤖 AI Agents: machine_sessions, machine_usage, machine_ai_actions
- 💳 Billing: user_credits, credit_transactions, stripe_customers, subscription_plans
- 📊 Projects: projects, user_machines, machine_snapshots
3. Set Up Environment Variables
# Frontend
cp .env.example .env
# Edit .env with your configuration
# Backend
cp backend/.env.example backend/.env
# Edit backend/.env with your configuration
Required Variables
Supabase (Required)
NEXT_PUBLIC_SUPABASE_URL=https://your-project.supabase.co
NEXT_PUBLIC_SUPABASE_ANON_KEY=your-anon-key-from-supabase-dashboard
SUPABASE_SERVICE_ROLE=your-service-role-key-from-supabase-dashboard
Security Keys (Required)
# Generate with: openssl rand -hex 32
ENCRYPTION_KEY=your-generated-32-byte-hex-string
CSRF_SECRET=your-generated-32-byte-hex-string
Google Search API (Required for web search)
GOOGLE_SEARCH_KEY=your-google-api-key
GOOGLE_SEARCH_CX=your-custom-search-engine-id
Get these from Google Cloud Console:
- Enable Custom Search API
- Create API key
- Create Custom Search Engine at programmablesearchengine.google.com
AI Provider Keys (Choose at least one)
# OpenAI
OPENAI_API_KEY=sk-...
# Anthropic
ANTHROPIC_API_KEY=sk-ant-...
# Azure OpenAI (Optional)
AZURE_OPENAI_ENDPOINT=https://your-endpoint.openai.azure.com/
AZURE_OPENAI_API_KEY=your-key
AZURE_OPENAI_DEPLOYMENT=your-deployment-name
AZURE_OPENAI_API_VERSION=2024-02-15-preview
Azure Container Instances (Optional - for cloud VM deployment)
AZURE_SUBSCRIPTION_ID=your
Frequently Asked Questions
What is open-computer-use?
open-computer-use is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by LLmHub-dev. The Open Framework for autonomous virtual computer agents at scale, fully open-source, safe, auditable, and production-ready. It has 316 GitHub stars.
Is open-computer-use safe to use?
Yes. open-computer-use passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
How do I install open-computer-use?
Clone the repository with "git clone https://github.com/LLmHub-dev/open-computer-use" and add it to your Claude Code skills directory (see the Installation section above).
What programming language is open-computer-use written in?
open-computer-use is primarily written in TypeScript. It is open-source under LLmHub-dev on GitHub, so you can review or fork the full source.
Are there alternatives to open-computer-use?
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh open-computer-use against similar tools.
Comments (0)
to leave a comment.
No comments yet. Be the first to share your thoughts!
Related Skills
An agentic skills framework & software development methodology that works.
234,966
20,863
Shell
AI Agentsaibrainstorming
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
228,318
35,019
JavaScript
AI Agentsai-agentsanthropic
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
196,003
59,243
TypeScript
MCP Serversaiapis
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
185,940
28,768
JavaScript
AI Agentsai-agentsanthropic
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.
120,031
19,897
Shell
AI Agents