Evaluates and scores code implementations using a standardized five-criteria quality framework.
The Speed-Run Judge skill provides a rigorous, objective framework for evaluating software implementations during Phase 4 of speed-run showdowns or any-percent workflows. It systematically analyzes code across five key dimensions: fitness for purpose, justified complexity, readability, robustness, and maintainability. By automating the comparison of different architectural approaches or LLM-generated variants, it ensures the selected winner meets high-quality standards while accounting for development efficiency metrics like hosted LLM calls and fix cycles.
주요 기능
01Direct comparison between multiple implementation variants (Showdown vs. Any%)
02Integration of speed-run efficiency metrics like fix cycles and generation time
03Automated gate checks for critical flaws and fitness deltas
0423 GitHub stars
05Transparent winner selection rationale with trade-off and token efficiency notes
06Standardized 5-criteria scoring framework for objective code evaluation
사용 사례
01Evaluating different architectural approaches to identify the most robust design
02Comparing multiple LLM-generated solutions for the same design specification
03Standardizing code quality audits during rapid prototyping and development sprints
