roborev has been designed to run code reviews triggered by git post-commit hooks (or enqueued by the roborev review CLI), so roborev spawns a non-interactive coding agent with a prompt to perform a review.
If a third party code reviewer does its own review, if the review outcome is not stored then that information is lost (for example many people use subagent code reviews from https://github.com/obra/superpowers — but these review outcomes are ephemeral). This creates an accountability problem. roborev serves essentially as a "review ledger" so that unaddressed reviews remain in the unaddressed ledger until explicitly addressed. So if a third party conducts a review, we would want to enable them to store the review findings in roborev so that we can come back later and see the findings (like we do in roborev now) as well as have a structured accountability loop to be able to verify that the findings were addressed.
roborev has been designed to run code reviews triggered by git post-commit hooks (or enqueued by the
roborev reviewCLI), so roborev spawns a non-interactive coding agent with a prompt to perform a review.If a third party code reviewer does its own review, if the review outcome is not stored then that information is lost (for example many people use subagent code reviews from https://github.com/obra/superpowers — but these review outcomes are ephemeral). This creates an accountability problem. roborev serves essentially as a "review ledger" so that unaddressed reviews remain in the unaddressed ledger until explicitly addressed. So if a third party conducts a review, we would want to enable them to store the review findings in roborev so that we can come back later and see the findings (like we do in roborev now) as well as have a structured accountability loop to be able to verify that the findings were addressed.