NYCU-Chung/my-claude-devteam ? reverse-engineered prompt
Reverse engineered prompt
Build me a Claude Code setup that feels like a real engineering team instead of one generic assistant.
I want a project I can drop into a repo that gives me a planner, engineer, refactor specialist, migration helper, designer, reviewer, security checker, debugger, database expert, onboarder, tool expert, and web researcher, each with their own job and permissions. It should also include the automation that keeps the team disciplined, like catching debugger statements, hardcoded secrets, stray console logs, and basic tool or config issues before I commit.
Make it easy to customize the agents, and make the whole thing work with clear instructions in the repo so I can use it in Claude Code right away. If you need to check the current Claude Code docs or any related setup details, go look them up online. I want the result to feel polished, practical, and ready for real day to day coding work.
Are you gonna build this?
make sure you review the code using coderabbit