Initial research release of SentinelProbe.
- API-based LLM probing
- Prompt batching, mutation, and replay
- Behavioral scoring (LOW / MED / HIGH)
- Indirect prompt injection testing
- Simulated RAG ingestion and persistence checks
- Structured logs and CSV evidence output