# hermes-max MCP server

> Provides coding agents with a suite of engineering tools including verification gates, repository maps, LSP, code graphs, and RAG.

- Repository: https://github.com/patrickbdevaney/hermes-max
- Page: https://hostedmcp.org/servers/patrickbdevaney/hermes-max
- Transport: —
- Stars: 1
- License: MIT

## FAQ

### What kind of models can I use with hermes-max?

It works with any OpenAI-compatible endpoint, including local inference via vLLM, llama.cpp, MLX, or Ollama, as well as hosted APIs.

### Is it compatible with other MCP clients besides Hermes Agent?

While only Hermes Agent has been tested end-to-end, the servers use standard MCP over HTTP and are expected to work with clients like Claude Code, Cursor, and VS Code.

### Does using hermes-max send my data to external providers?

Prompts and code context are sent to your configured hosted providers; however, if you only configure local endpoints, no model traffic leaves your machine.

### How does the verification gate work?

It uses a pipeline of linting, type-checking, and testing as a completion gate, including property and mutation testing.