llm-security-prompt-injection
The calm guardian for AI, agent, and MCP code — a zero-dependency, OWASP-anchored AI security toolkit.
Static security analysis for AI, agent and MCP code. Detects security-sensitive prompt/context, tool, memory, model-output and credential patterns. Local, deterministic and CI/SARIF-ready.
Bitmask-based LLM security firewall with policy-driven logits filtering using GPU-accelerated Aho-Corasick pattern matching
Defense-in-depth input safety for LLMs — perplexity gate + FAISS + ModernBERT + LoRA + Llama Guard 3, behind a deterministic policy gate. 99.88% accuracy, 99.47% jailbreak recall, calibrated confidence, ONNX-optimized. Live demo on HF Spaces.
