ReasonGate- An explainable gate that blocks LLM prompt injection

ReasonGate is an open-source security tool that blocks prompt injection attacks on AI applications before they reach the language model. Prompt injection, the top vulnerability on the OWASP LLM Top 10, tricks AI models into following hidden malicious instructions. Unlike typical security tools that give a vague yes/no score, ReasonGate explains exactly why it blocked something, making decisions auditable for security teams and regulators. It catches 96% of known attacks with only 0.3% false positives.