# OpenMake MCP server

> An open-source AI workspace that orchestrates local and open-weight models, autonomous agents, and MCP tools through a self-hosted runtime.

- Repository: https://github.com/openmake/openmake_llm
- Page: https://hostedmcp.org/servers/openmake/openmake-llm
- Transport: —
- Stars: 44
- License: Apache-2.0

## FAQ

### What is the primary purpose of OpenMake?

OpenMake is designed to coordinate models, tools, and agents to perform general AI work on infrastructure you control.

### Can I use my own API keys with OpenMake?

Yes, you can bring your own keys (BYOK) for various providers, and they are stored using AES-256-GCM encryption.

### Does OpenMake support local models?

Yes, it supports local models through vLLM, Ollama, or any OpenAI-compatible endpoint.

### How does OpenMake handle agent tasks?

Agents can maintain task state, use tools under various approval policies, execute code in Docker sandboxes, and browse the web.
