This MCP server provides a fully-equipped Ubuntu sandbox for any LLM, allowing code execution, document creation, web browsing, and the use of autonomous sub-agents in isolated Docker containers.
From the registry: Give any LLM its own computer — Docker sandboxes with bash, browser, docs, and sub-agents

$ git clone https://github.com/Yambr/open-computer-use.git && cd open-computer-use && cp .env.example .env && docker compose up --buildPlease install the `open-computer-use` MCP server into my current AI client (that's you).
Required prerequisites (do these first if not already done):
- **Docker** — Requires Docker for running the workspace and sandbox containers.
- **OpenAI-compatible LLM API key** — Any OpenAI-compatible provider (OpenAI, OpenRouter, etc.).
- **Clone and start** — Clone the repo, configure .env, and bring up the docker-compose stack. Run: `git clone https://github.com/Yambr/open-computer-use.git && cd open-computer-use && cp .env.example .env && docker compose up --build`
Canonical MCP server config (HTTP transport):
- url: `http://localhost:8081/mcp`
- headers:
- `Authorization`: `Bearer <your-mcp-api-key>`
- `X-Chat-Id`: `<chat-id>`
Note: Self-hosted sandbox that gives LLMs a Docker-isolated Ubuntu environment. MCP_API_KEY header required in production.
Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.OPENAI_API_KEYrequiredLLM API key (any OpenAI-compatible)MCP_API_KEYrequiredBearer token for MCP endpointVISION_API_KEYrequiredVision API key (for describe-image)ANTHROPIC_AUTH_TOKENrequiredAnthropic key (for Claude Code sub-agent)An MCP Server for computer-use in Windows OS