humane proxy for Claude Code

io.github.vishisht16/humane-proxy

AI safety middleware — detects self-harm and criminal intent in LLM prompts.

client:Claude Code transport:stdio runtime:pypi

Install humane proxy in Claude Code

run in your project directory

claude mcp add humane-proxy -e OPENAI_API_KEY=<OPENAI_API_KEY> -e GROQ_API_KEY=<GROQ_API_KEY> -- uvx humane-proxy

Adds it for this project only. Append --scope user to make it available everywhere. Claude Code docs