Humane Proxy
AI safety middleware that detects self-harm and criminal intent in LLM prompts.
Humane Proxy acts as an AI safety layer that screens LLM prompts for self-harm indicators and criminal intent before they reach downstream models. Using multi-stage detection including optional LLamaGuard and OpenAI reasoning, the server helps developers build safer AI applications.
Source
Repository: https://github.com/vishisht16/humane-proxy
Maintain Humane Proxy?
Let people know it's listed here — add the badge (live metrics, light/dark aware) or a plain link to your README or docs.
[](https://getagentictools.com/mcp/vishisht16-humane-proxy?ref=badge) Related mcp
Whoiser Provides a lightweight server for performing WHOIS lookups using the whoiser library, enabling retrieval of domain registration an
Currents Test Results Provides a bridge to Currents test results platform, enabling AI to analyze failing tests, optimize test suites, and troubleshoot
Attestix AI agent identity, W3C credentials, and EU AI Act compliance tools.
Code Graph Context Builds rich code graphs to provide deep contextual understanding of TypeScript codebases
arifOS Constitutional AI governance kernel that enforces 13 safety rules between language models and tools.
Apache APISIX Bridge LLMs with the APISIX Admin API to manage and analyze API gateway information.