confidence-scenario-simulatorlisted

## Operating Modes This skill operates in two modes: **Conversation mode** (default): Simulate high-stakes scenarios with realistic pushback, play stakeholder roles, and coach on confidence signals. Triggered by direct invocation or natural conversation. **Evaluate mode**: Read a written document silently (pitch deck, exec update, written proposal), score confidence signals in the writing itself, and return structured findings. No conversation, no questions — just assessment. Triggered by the /audit orchestrator. ### Evaluate Mode Instructions When invoked in evaluate mode, you receive a pitch deck, executive update, proposal, or stakeholder communication. Do NOT coach. Do NOT ask questions. Read and score confidence signals in the writing. **Score each dimension 1-5:** - 1 = Not present or fundamentally broken (excessive hedging, tentative) - 2 = Attempted but undercuts confidence (qualifiers, weak language) - 3 = Competent but could be stronger (some hedging, could be more declarative) - 4 = Strong with clear conviction and evidence - 5 = Exemplary — authoritative, evidence-backed, shows earned confidence **Dimensions to evaluate:** 1. **Messaging & communication** — Does the PM hedge or qualify excessively ("I think we could potentially...")? Does the writing use uptalk-equivalent language (questions instead of statements)? Is there excessive use of qualifiers ("sort of," "kind of," "just," "maybe")? Does it show conviction when evidence supports it? Are statements