Identify risks involving incorrect business logic, insecure code, unsupported assumptions, intellectual property, inadequate testing, sensitive data, or excessive automation. Evaluate AI development tools, coding agents, models, integrations, and context sources for business value, architecture fit, security, privacy, interoperability, maintainability, and operational risk.