Interview Prep Coach

Generates 5 likely questions for the role + STAR-format answer scaffolding.

## Interview Prep: {role} at {company}

**Inputs**
- Role: {role}
- Company: {company}

### Output

5 questions, mix: 2 behavioral, 2 technical/role-specific, 1 curveball.

For each:

**Q[N] — [Behavioral / Technical / Curveball]**
- **Question**: phrased in real interviewer voice.
- **What they're really evaluating**: 2-3 specific, ratable signals (e.g., "scope of ownership", "tradeoff articulation under ambiguity").
- **STAR scaffold**:
  - **Situation**: [scope, team size, stakes — 1-2 sentences]
  - **Task**: [your specific responsibility]
  - **Action**: [2-3 decisions + the tradeoff weighed]
  - **Result**: [outcome with a real metric, plus what you'd do differently]

Tailor topic areas to {role} and {company}'s known interview loop where possible.

### Rules

- ✅ Real interviewer language; signals are ratable (not "good communicator")
- ✅ Curveball reflects {company} culture (values prompt at Amazon, disagree-and-commit at Meta)
- ❌ "Tell me about yourself" softballs; vague signals; generic STAR placeholders
- ❌ Filler like "passionate", "team player", "hard-working" anywhere in the scaffold
- If {role}/{company} is too broad, write `Need: [specific level/team]` instead of guessing