Orgo MCP Server
Enables AI agents to control virtual computers through Orgo, supporting computer management, screen actions, shell commands, file operations, streaming, and AI completions.
README
Orgo MCP Server
An MCP (Model Context Protocol) server that gives AI agents the ability to control virtual computers through Orgo.
Get Started in 60 Seconds
+-------------------------------------------------------------+
| STEP 1: Get Key > STEP 2: Configure > STEP 3: Use! |
+-------------------------------------------------------------+
Step 1: Get Your API Key
| Action | Link |
|---|---|
| New user? | Sign up at orgo.ai |
| Existing user? | Go to Dashboard |
Where to find your key:
+----------------------------------------------------------+
| orgo.ai Dashboard |
+----------------------------------------------------------+
| |
| Settings > API Keys > [ Copy Key ] |
| |
| Your key looks like: sk_live_abc123... |
| ^^^^^^^^ |
| Always starts with sk_live_ |
| |
+----------------------------------------------------------+
Tip: Keep your API key private - never commit it to git!
Step 2: Configure Your Client
Choose your setup method:
<details open> <summary><strong>Claude Code (CLI) - Recommended</strong></summary>
Run this single command:
claude mcp add --transport http orgo https://orgo-mcp.onrender.com/mcp \
--header "X-Orgo-API-Key: YOUR_API_KEY"
# ^^^^^^^^^^^^^
# Replace this with your actual key!
Example with a real key format:
claude mcp add --transport http orgo https://orgo-mcp.onrender.com/mcp \
--header "X-Orgo-API-Key: sk_live_abc123xyz789"
</details>
<details> <summary><strong>Claude Desktop (macOS)</strong></summary>
1. Open your config file:
open ~/Library/Application\ Support/Claude/claude_desktop_config.json
2. Add this configuration:
{
"mcpServers": {
"orgo": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://orgo-mcp.onrender.com/mcp",
"--header", "X-Orgo-API-Key:sk_live_YOUR_KEY_HERE"]
}
}
}
Important: No space after the colon in
X-Orgo-API-Key:sk_live_...
3. Restart Claude Desktop
</details>
<details> <summary><strong>Claude Desktop (Windows)</strong></summary>
1. Open your config file:
%APPDATA%\Claude\claude_desktop_config.json
2. Add this configuration:
{
"mcpServers": {
"orgo": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://orgo-mcp.onrender.com/mcp",
"--header", "X-Orgo-API-Key:sk_live_YOUR_KEY_HERE"]
}
}
}
Important: No space after the colon in
X-Orgo-API-Key:sk_live_...
3. Restart Claude Desktop
</details>
<details> <summary><strong>Team Project (.mcp.json)</strong></summary>
For sharing with your team without exposing keys:
1. Create .mcp.json in your project root:
{
"mcpServers": {
"orgo": {
"type": "http",
"url": "https://orgo-mcp.onrender.com/mcp",
"headers": {
"X-Orgo-API-Key": "${ORGO_API_KEY}"
}
}
}
}
2. Each team member sets their own key:
# Add to ~/.bashrc or ~/.zshrc for persistence
export ORGO_API_KEY="sk_live_your_personal_key"
</details>
Step 3: Start Using!
Try these commands in Claude:
"Create a Linux computer with 4GB RAM"
"Take a screenshot of my computer"
"Run 'ls -la' on the computer"
"Type 'hello world' and press Enter"
Troubleshooting API Key Issues
<details open> <summary><strong>Common Problems and Quick Fixes</strong></summary>
| Error Message | What is Wrong | How to Fix |
|---|---|---|
Invalid API key |
Key format is wrong | Make sure it starts with sk_live_ |
X-Orgo-API-Key header required |
Key not passed correctly | Check for extra spaces in header |
401 Unauthorized |
Key is expired or invalid | Generate a new key at orgo.ai |
Quick validation checklist:
[ ] Key starts with: sk_live_
[ ] No extra spaces before/after the key
[ ] Key copied completely (no truncation)
[ ] Quotes are straight " not curly
</details>
Features (34 tools)
| Category | Tools |
|---|---|
| Projects | orgo_list_projects, orgo_create_project, orgo_get_project, orgo_delete_project, orgo_start_project, orgo_stop_project, orgo_restart_project |
| Computers | orgo_list_computers, orgo_create_computer, orgo_get_computer, orgo_start_computer, orgo_stop_computer, orgo_restart_computer, orgo_delete_computer |
| Actions | orgo_screenshot, orgo_click, orgo_double_click, orgo_type, orgo_key, orgo_scroll, orgo_drag, orgo_wait |
| Shell | orgo_bash, orgo_exec |
| Files | orgo_list_files, orgo_upload_file, orgo_export_file, orgo_download_file, orgo_delete_file |
| Streaming | orgo_start_stream, orgo_stream_status, orgo_stop_stream |
| AI | orgo_list_ai_models, orgo_ai_completion |
Self-Hosting Options
Option 1: Local Development (stdio)
git clone https://github.com/nickvasilescu/orgo-mcp.git
cd orgo-mcp
pip install -e .
export ORGO_API_KEY="your_key"
python orgo_mcp.py
Add to claude_desktop_config.json:
{
"mcpServers": {
"orgo": {
"command": "python3",
"args": ["/path/to/orgo-mcp/orgo_mcp.py"],
"env": {"ORGO_API_KEY": "your_key"}
}
}
}
Option 2: Local HTTP Server
pip install -e .
MCP_TRANSPORT=http python orgo_mcp.py
Server available at http://localhost:8000/mcp
Option 3: Docker
# Build
docker build -t orgo-mcp .
# Run
docker run -p 8000:8000 orgo-mcp
# Or use docker-compose
docker-compose up
Option 4: Deploy to Render.com
- Fork this repository
- Go to Render Dashboard
- Click "New Blueprint Instance"
- Connect your GitHub repo
- Deploy!
Your server will be at: https://orgo-mcp-xxxx.onrender.com
Option 5: Deploy to Fly.io
# Install flyctl
curl -L https://fly.io/install.sh | sh
# Login and deploy
fly auth login
fly launch --no-deploy
fly deploy
Your server will be at: https://orgo-mcp.fly.dev
Environment Variables
| Variable | Default | Description |
|---|---|---|
MCP_TRANSPORT |
stdio |
Transport mode: stdio or http |
MCP_HOST |
0.0.0.0 |
HTTP bind address |
MCP_PORT / PORT |
8000 |
HTTP port |
ORGO_API_KEY |
- | API key (required for stdio transport) |
CORS_ORIGINS |
* |
Allowed origins (comma-separated) |
Usage Examples
Project Management
"Create a new project called 'qa-automation'"
"List my Orgo projects"
"Start all computers in project proj_123"
"Delete project proj_123"
Computer Management
"Create a new Linux computer called 'dev-box' with 4GB RAM"
"Get details for computer abc123"
"Start computer abc123"
"Stop computer abc123"
Screen Actions
"Take a screenshot of computer abc123"
"Click at coordinates (500, 300)"
"Type 'hello world'"
"Press Enter"
"Scroll down"
Shell Commands
"Run 'ls -la' on computer abc123"
"Execute Python code: print('hello')"
File Operations
"List files on computer abc123"
"Export file ~/Documents/report.pdf from computer abc123"
"Upload file to computer abc123"
Streaming
"Start streaming computer abc123 to Twitch"
"Check stream status for computer abc123"
"Stop streaming computer abc123"
AI Completion
"List available AI models"
"Run GPT-4 completion: 'Explain quantum computing'"
Additional Troubleshooting
"Connection refused" Error
- Cloud: Check
https://orgo-mcp.onrender.com/health - Local: Ensure server is running on correct port
- Check firewall/proxy settings
Claude Desktop Not Connecting
- Restart Claude Desktop after config changes
- Check config file syntax (valid JSON)
- View logs: Help > Troubleshooting > Open Logs
Tools Not Appearing
- Wait 10-30 seconds after connection
- Check server logs for errors
- Verify MCP server is listed in Claude settings
Architecture
Cloud Deployment:
Client (Claude) --> HTTPS --> Cloud Server --> Orgo API
X-Orgo-API-Key header
Local Deployment:
Client (Claude) --> stdio --> orgo_mcp.py --> Orgo API
ORGO_API_KEY env var
Project Structure
orgo-mcp/
+-- orgo_mcp.py # MCP server (34 tools, dual transport)
+-- pyproject.toml # Package configuration
+-- Dockerfile # Production container
+-- docker-compose.yml # Local development
+-- render.yaml # Render.com deployment
+-- fly.toml # Fly.io deployment
+-- .dockerignore # Docker build exclusions
+-- .env.example # Environment template
+-- README.md
+-- LICENSE
Development
# Install dev dependencies
pip install -e ".[dev]"
# Format code
black orgo_mcp.py
# Lint
ruff check orgo_mcp.py
# Run tests
pytest
# Test HTTP transport locally
MCP_TRANSPORT=http python orgo_mcp.py
# Test health endpoint
curl http://localhost:8000/health
# Test MCP endpoint
curl -X POST http://localhost:8000/mcp \
-H "Content-Type: application/json" \
-H "X-Orgo-API-Key: your_key" \
-d '{"jsonrpc":"2.0","method":"tools/list","id":1}'
License
MIT License - see LICENSE
Credits
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。