llm-protection

On-prem detection & anonymization of PII and sensitive data for LLM inputs — hybrid regex + NER + local LLM

🛡️ Protect LLM applications with PromptShields, a robust security framework designed to prevent prompt injection, jailbreaks, and data leakage.

The open-source, extensible LLM gateway and agentic MCP proxy. Build, audit, and route prompts to 100+ LLMs with unified telemetry and access controls