Quantify variance in stochastic skill-review scores across parallel runs.
npx skills add https://github.com/Alexmacapple/alex-claude-skill --skill eval-robuste
Please help me install this Agent Skill. Skill: eval-robuste Source: https://github.com/Alexmacapple/alex-claude-skill/tree/main/eval-robuste Command: npx skills add https://github.com/Alexmacapple/alex-claude-skill --skill eval-robuste
Measures the variance in stochastic skill evaluations (skill-review) by running N parallel instances, aggregating results deterministically, and producing a clear verdict against a reference to decide if a perceived change is noise or meaningful.
Invoke the robust evaluation on a target skill to quantify evaluator noise and determine if observed changes indicate a true regression.
High-intent search queries and answers about installing and using this skill.
To check if evaluation score changes are noise, run N parallel skill-review instances to quantify variance. This deterministic aggregation computes median and standard deviation, producing a clear verdict on whether a change is meaningful or just noise.
The best way to establish a baseline sigma for stochastic skill evaluations is by running multiple parallel passes. The process aggregates raw scores deterministically and outputs a baseline JSON file for reliable future regression comparisons.
To quantify variance when skill-review results fluctuate across passes, launch N parallel evaluations concurrently. A dedicated aggregator script computes the standard deviation and confidence-like interval, determining if the score oscillation indicates true regression.
Yes, parallel skill evaluation runs are enforced between N equals 2 and 10. You can execute larger runs by explicitly using the force flag, ensuring you acknowledge the increased computational overhead for extensive variance testing.
Robust skill evaluation outputs a Markdown report and a baseline JSON file. The Markdown report details the deterministic verdict of BRUIT, REGRESSION, or AMELIORATION, while the JSON provides the baseline sigma for future regression decisions.
Deterministic aggregation handles evaluator noise by delegating raw scores from parallel skill-review runs to a dedicated script. It computes the median, standard deviation, and confidence-like interval to deliver a clear verdict against an archived reference.
xurikValidate skill evaluation quality metrics and regression readiness.
AlexandreLabAutomate end-to-end web performance benchmarking with baseline comparison and JSON/markdown reports.
lohasleApply descriptive statistics, trend analysis, outlier detection, and hypothesis testing to datasets.
zz8011Monitors deployed pages for console errors, performance issues, and page failures against baselines.
