Responsible AI Plan from Security Plan
Independent PiSkill directory guide. The original prompt remains hosted by Microsoft HVE Core.
What does this prompt do?
Extends an existing security plan with model-specific harms, misuse, data, evaluation, transparency, monitoring, and governance requirements.
Primary use case
Add responsible-AI controls to an established security plan.
Expected output
An integrated set of responsible-AI risks, controls, and validation work.
Inputs or variables
- security plan
- model behavior
- users
Related prompts
Responsible AI Evidence Capture
Captures responsible-AI requirements, stakeholders, harms, controls, evidence, ownership, and unresolved questions in a consistent planning artifact.
Security Plan from PRD
Derives a security plan from product requirements by mapping assets, data, trust boundaries, abuse cases, controls, validation, and operational ownership.
Incident Response Plan
Structures an incident response workflow around evidence collection, containment, recovery and follow-up actions.
LLM Security Review
Reviews an LLM or agent solution for prompt injection, unsafe tool access, data exposure, trust-boundary failures, weak validation, and operational controls.
Responsible AI Plan from PRD
Turns a product requirements document into a responsible-AI plan covering likely harms, mitigations, validation, monitoring, governance, and accountable owners.
Secure Supply Chain Plan from PRD
Derives software-supply-chain controls from product requirements, covering dependencies, provenance, builds, artifacts, releases, ownership, and verification.