Show HN: Retroguard – Verifiably secure AI guardrails

Category: security

Tags: ai, security, openai-proxy

Score: 4.3/10 (Innovation: 4, Technical: 4, Documentation: 2, Utility: 5)

Retroguard provides a drop-in proxy for OpenAI's API that adds verifiably secure guardrails to AI interactions, aiming to prevent sensitive data leakage. The project is interesting for its focus on privacy verification, but lacks substantial open-source artifacts or documentation.

Target audience: backend devs, devops

Repository: https://retroguard.ai

View on Hacker News