What problem does it solve? Multi-option decisions with competing criteria often get decided by the loudest stakeholder or by numbers arranged after the fact to justify a favourite. This Skill makes trade-offs explicit: it structures criteria, derives weights (including AHP pairwise comparisons with a consistency check), scores options on a common scale, and quantifies how robust the winner is to weight and score changes. ## Core Features & Use Cases - Weighted decision matrix: Scores 3+ options against 3-7 criteria with direction-aware normalisation (min-max or ratio), weighted totals, ranking, and the margin between the top two. - AHP weight derivation: Builds Saaty's reciprocal pairwise comparison matrix, computes the principal eigenvector as weights, and reports lambda_max, CI, RI, and CR with a consistency verdict (CR <= 0.10 gate). - Sensitivity analysis: Reports per-criterion break-even weights, the single score cell whose smallest change flips the leader, and a leave-one-out rank-reversal check. - Use Case: Choosing one of four R&D programmes for a budget: define criteria (strategic fit, NPV, time to revenue, risk), derive AHP weights, and get a ranking plus the finding that a EUR 2-3 m NPV error would flip the winner. ## Quick Start Ask the assistant to build a weighted decision matrix for your options and criteria, deriving the weights with AHP pairwise comparisons and reporting how robust the winner is.
