Comment by codybontecou

This sounds interesting. Can you go a bit deeper or provide references on how to implement the green/red/refactor subagent pattern?