| name | practice-scenario |
| description | Creates fictional GRC scenarios and grades learner reasoning, assumptions, and tradeoffs. USE WHEN learner wants judgment practice, interview-style rehearsal, applied case work, or a safe way to try decisions before explanation. NOT FOR analyzing a real vendor, audit, control, policy, or programme. |
practice-scenario
What
Generate realistic but fictional situations for the learner to respond to. Grade the reasoning, not just the final answer.
When
- The learner wants judgment practice.
- The learner asks for a mock scenario, case, or interview-style prompt.
- The learner needs to practise tradeoffs after learning a concept.
- The learner wants a safe substitute for a live operational request.
Not For
- Reviewing a real vendor, control, audit, policy, or programme.
- Coaching by questions before a scenario is framed. Use
socratic-coach.
- Building a project artefact. Use
lab-builder.
Inputs
- Target concept or skill.
- Learner level and timebox.
- Optional corpus ideas when the scenario needs citation grounding.
Steps
- Pick a fictional organization, role, constraint, and ambiguous signal.
- State only the facts needed for the exercise.
- Ask the learner what they would do first and why.
- Let them answer before giving the stronger answer.
- Grade reasoning, assumptions, tradeoffs, and boundary awareness.
- Ask what principle transfers to another setting.
Validation
- The case is fictional and does not require sensitive details.
- Feedback names reasoning quality, not just correctness.
- The learner leaves with one transferable principle.
Gotchas
- If the learner starts describing their real case, abstract it before continuing.
- If the scenario has a single obvious answer, add an ambiguous constraint or tradeoff.
- If the learner is too new for the case, teach the prerequisite with
concept-tutor first.
Failure Modes
- Operational leak: remove real organizations, vendors, evidence, controls, and deadlines.
- Answer-first teaching: require the learner's attempt before giving model reasoning.
- Overloaded prompt: keep the scenario small enough for one response.
Examples
- User asks for SOC 2 practice -> Create a fictional startup with messy access review evidence and ask what signal they would inspect first.
- User wants interview prep -> Give a toy TPRM tradeoff case, grade assumptions, then ask for a revised answer.
- User asks "Should we approve this vendor?" -> Refuse the live decision and convert it into a fictional vendor reasoning scenario.