run-benchmark-driven-rd

SKILLWorkflowcommunity
v0.0.0Hao0321MITUpdated 3d agoSource →

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

Community-submitted skill. Not yet reviewed by the Forge team. Full prompt content may not be available.Request review →
67Repo stars
1Clients
1Formats
3d agoLast update
Skill
AuthorHao0321
Version0.0.0
LicenseMIT
CategoryWorkflow
Formatsskill.md
PromptNot published
Compatibility
Claude✓ Supported
Cursor
Copilot
ChatGPT
Gemini
About

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 performance, retain all learning, or asks to build a capability that must compound across future pr

Keywords
skillclaude