Skill

run-benchmark-driven-rd

Turn ambitious or uncertain engineering goals into a falsifiable, benchmark-gated R&D system with calibrated evaluation tools, architecture gates, external-change safety, and reusable experiment memory. Use when the user says 做到最好, 全面重構, 架構最優, 超越競品, 自己研發, 研發系統, benchmark, R&D, improve recognition or

Claim this listing

Connect your GitHub to prove you own or maintain this listing. We verify repo access automatically — most publishers are confirmed in seconds.

1Connect GitHub
2Submit your claim
3Auto-verified, or reviewed within 48h