LLM testing

Step-by-step playbook to run a lightweight internal AI red team: scope, prioritized test cases, ready-to-run prompts, severity scoring, reporting, and remediation.