Responsible AI Development

Suite of tools providing model and data exploration, assessment interfaces and libraries for understanding AI systems. Enables developers to develop and monitor AI more responsibly with better data-driven actions. MIT licensed.

Security and compliance toolkit for autonomous agents with policy enforcement, zero-trust identity, and sandboxed execution to reduce operational risk. MIT licensed.

Open database of real-world AI agent security incidents, with each sourced record flagged for confirmed harm, AI involvement and kind, plus JSON/CSV exports. CC BY 4.0 licensed.