Show HN: See what your employees are prompting LLMs (without network proxies)

Category: security

Tags: ai-security, compliance, agent-monitoring

Score: 6.0/10 (Innovation: 6, Technical: 5, Documentation: 3, Utility: 7)

Privent is a security layer for agentic AI that monitors LLM prompts within execution graphs (like LangGraph, CrewAI) without requiring network proxies. It's interesting because it addresses emerging compliance needs (GDPR, HIPAA, EU AI Act) for AI agents by providing real-time data protection at the agent runtime level.

Target audience: devops, ai-engineers, security-engineers

Repository: https://www.privent.ai/

View on Hacker News