What problem does it solve? Complex, hard-to-reverse decisions often suffer from single-perspective thinking, hidden assumptions, and premature convergence. This Skill structures deliberation as a four-lens expert panel—two real public figures plus domain-expert and omniscient-agent archetypes—that debate, cross-examine, and converge on a recommendation with explicit uncertainty tracking. ## Core Features & Use Cases - Four-Lens Deliberation Framework: Runs fixed multi-round dialogues (initial positions, cross-examination, revisions, final statements) with mandatory inference-confidence tags and evidence anchors for real-person viewpoints. - Controlled Execution Profiles: Choose micro (2 roles, 5 sections), lean, classic (4 roles, 7 sections), or deep (6 rounds, 9 sections with stress tests and contingency planning) based on decision complexity. - Blind-Spot Gated Audit: An optional overlay routes the candidate decision to a heterogeneous cross-vendor reviewer, applies a materiality gate, and triggers a single-shot re-entry only when a material blind spot is found. - Use Case: When deciding whether to split a monolith into microservices, the Skill selects two relevant real engineers with opposing pressures, simulates their debate, and delivers a synthesis with preconditions, early-warning indicators, and an uncertainty ledger. ## Quick Start Ask the agent to use simulate-elite-experts to analyze your decision question, optionally specifying a profile such as micro or deep.
