Problem / Use case
I realize this is sort of against the spirit of Ralph loops overall, but the idea here is when converting the prd to tasks, an option can be used to have the AI place logical checkpoints in the plan that would include its own end to end testing specific to your app, and then generating UAT test cases for the user to run to validate/try their project out along the way. Then you could have the option for running the ralph loops until the next checkpoint (with a max iterations limit as well just in case).
Would allow for more steering of the project by the user as they built it out.
Proposed solution
Alternatives considered
No response
Additional context
No response
Problem / Use case
I realize this is sort of against the spirit of Ralph loops overall, but the idea here is when converting the prd to tasks, an option can be used to have the AI place logical checkpoints in the plan that would include its own end to end testing specific to your app, and then generating UAT test cases for the user to run to validate/try their project out along the way. Then you could have the option for running the ralph loops until the next checkpoint (with a max iterations limit as well just in case).
Would allow for more steering of the project by the user as they built it out.
Proposed solution
Alternatives considered
No response
Additional context
No response