What problem does it solve? Translating a mathematical optimization problem into correct CVXPY code is error-prone: curvature rules (DCP/DPP/DGP), shape semantics, solver selection, and status handling all cause silent failures or wrong results. This Skill guides writing, reviewing, debugging, and validating CVXPY models so the code matches the intended mathematics. ## Core Features & Use Cases - Disciplined model construction: Enforces correct use of Variable, Parameter, Expression, Constraint, Objective, and Problem, including DCP/DPP ruleset checks before solving. - Solver and status discipline: Guides solver selection from installed capabilities, branches on OPTIMAL, OPTIMAL_INACCURATE, INFEASIBLE, and UNBOUNDED, and blocks reading values before status gates. - Independent validation: Requires residual, integrality, objective, and dual/KKT checks at justified tolerances rather than trusting solver output. - Use Case: You need a bounded least-squares model whose coefficients change between runs. The Skill produces a Parameter-based DPP-compliant formulation, verifies is_dcp(), selects an installed solver, and validates primal feasibility before returning results. ## Quick Start Ask the AI to write a CVXPY model for your optimization problem, including DCP checks, solver selection, status handling, and residual validation.
