What problem does it solve? Hackathon teams routinely converge on the same obvious ideas, pick concepts they cannot build in the timebox, or choose projects that score poorly on the actual judging rubric. This Skill replaces unstructured brainstorming with a timed, evidence-driven procedure that produces a differentiated, buildable, rubric-aligned idea. ## Core Features & Use Cases - Consensus Banning: Simulates the average team's brainstorm, lists 12-15 predictable ideas, and bans them so divergence rounds cannot regress to consensus. - Evidence-Grounded Divergence: Requires web research on user pain, usable APIs and datasets, and prior winners before any candidate idea is accepted; every candidate must cite a verified finding. - Feasibility and Judge Simulation: Scores survivors on demo-impact versus build-cost, runs a deployability battery (rails, operator, zero-training, cold-start, unit economics, zero-bars), and outputs a ranked IDEA-BRIEF. - Use Case: A four-person team at a 36-hour sponsor-judged hackathon runs the gauntlet in 90 minutes, kills the obvious chatbot ideas, and commits to an offline-first IVR solution with a verified delivery rail and a sponsor-prize portfolio. ## Quick Start Run the ideation gauntlet on this hackathon problem statement and judging criteria to produce an IDEA-BRIEF with a ranked shortlist of buildable ideas.
