A rigorous, standards-aligned, learning-first assessment covering LLM application security, prompt injection, RAG security, agent security, model supply chain, MLOps, AI SDLC, privacy, governance, red teaming, incident response, vendor risk, and secure architecture. This is a readiness benchmark and educational resource, not an accredited certification.
agent security
A research agent can delegate tasks to a browsing agent, a code agent, and a reporting agent. Each sub-agent has different tools and permissions.