solution-iteration

Solution-iteration provides structured comparison by documenting each approach and running independent evaluations through diverse LLM judge personas—Pragmatist, Innovator, User Advocate, Business Strategist, Technical Architect, and Security Auditor—to surface trade-offs across usability, performance, business value, accessibility, and security.

Yes. This skill applies LLM judges to product, design, and engineering decisions where teams balance multiple constraints. Each judge persona evaluates independently, then results are synthesized into a recommendation with documented trade-offs to guide implementation.

Solution-iteration accelerates convergence through rapid evaluation cycles: describe approaches, run judge evaluations, synthesize insights, and refine using CLI commands. This workflow supports A/B/n testing, progressive refinement, and design space exploration patterns.

The skill generates decision records and implementation plans that capture each approach's strengths and limitations as judged by predefined personas, creating an audit trail of reasoning that justifies the final recommendation.

No. The skill uses built-in judge personas with inherent evaluation perspectives—technical, business, user-focused, security, and innovation angles—eliminating the need to manually construct evaluation rubrics.