Standardized benchmark for testing AI safety scanners. Run your scanner, get a score. From 15 Research Lab.

Comprehensive AI security payloads and wordlists. The SecLists of AI. From 15 Research Lab.

Map the attack surface of your AI agents. Enumerates tools, capabilities, and security gaps. SARIF output for GitHub Security tab.

Behavioral fingerprinting for AI agents. Build profiles, detect drift, identify compromise.

Bidirectional mapping between MITRE ATT&CK and AI alignment failure modes. The rosetta stone between cybersecurity and AI safety.

Black box recorder for AI agents. Reconstruct decisions, detect anomalies, verify audit chains.

Can AI systems detect when they're being evaluated? Research paper and reference implementation exploring the Hawthorne Effect for AI.