Add Sentinel AI

[MaxwellCalkin]

Adds Sentinel AI — a real-time safety guardrails library for LLM applications.

What it does

Sentinel AI scans LLM prompts and outputs for security threats and safety violations:

• 8 scanners: prompt injection, PII detection, harmful content, hallucination, toxicity, blocked terms, tool-use validation, structured output validation
• Multilingual injection detection: 12 languages + cross-lingual injection
• MCP Safety Proxy: transparent safety layer for any MCP server
• Sub-millisecond latency (~0.05ms average scan time)
• 530 benchmark cases, 100% accuracy (0 false positives, 0 false negatives)

Why it fits this list

Sentinel AI is a static/runtime analysis tool focused on security scanning for LLM applications — detecting prompt injection attacks, PII leakage, and other safety violations before they reach production. Tagged with python and security.

• License: Apache 2.0
• Language: Python
• Type: CLI + Library
• Live demo