Identify and retry discarded experiments after architectural changes.
npx skills add https://github.com/SteeZyT33/perf-lab-plugin --skill replay
Please help me install this Agent Skill. Skill: replay Source: https://github.com/SteeZyT33/perf-lab-plugin/tree/main/skills/replay Command: npx skills add https://github.com/SteeZyT33/perf-lab-plugin --skill replay
This Skill helps re-evaluate experiments that were previously discarded, especially after architectural changes, to identify potential new optimizations.
Use the replay skill to find and retry old experiments that might work now.
High-intent search queries and answers about installing and using this skill.
Re-evaluate discarded optimization experiments by analyzing experiment history against recent changelogs to determine viability, identifying which previously failed hypotheses might now succeed and positively impact performance due to specific architectural modifications.
Identify previously failed code changes for re-evaluation by automatically scanning experiment history to find optimization hypotheses marked as discarded prior to the last architectural change, then assessing their potential impact in the evolving codebase.
Architectural modification impacts regression testing hypothesis viability by altering the underlying codebase structure, which can resolve previous bottlenecks and allow previously discarded performance tuning experiments to succeed and deliver new optimizations.
You can selectively retry the most promising discarded performance tuning experiments by analyzing experiment history and changelogs to filter out hypotheses with the highest potential impact, facilitating targeted re-evaluation without testing every discarded change.
You should not re-evaluate old optimization hypotheses when the architectural changes are unrelated to the original failure points, or when the experiment history lacks sufficient context to accurately assess the potential impact of retrying the discarded code changes.
dangelesDesign experimental protocols with success criteria, controls, and prioritized experiments.
JaCha00Automates evaluator-gated performance tuning with durable artifacts in .omx/goals/performance/
wachawoDiagnose software bugs and performance regressions through a six-phase methodology.
ZateRoute codebases to vulnerability-pattern detection skills for security scanning.
