Hi there, thanks for making this plugin! I've found it super useful so far. A few weeks ago while I was first getting started with it, I had the idea that in addition to tech spec and PRD modes, it could be useful to apply the same iterative adversarial model to code reviews. I actually used the plugin to first create a spec for this and then implement it here: main...bgentry:adversarial-spec:bg/code-review
It worked pretty well the times I've tried it so far! Getting the agents to argue about which code review issues are legit, worth fixing, and how to fix them has been a good way to quickly improve complicated PRs. I definitely haven't put a lot of time into getting it ready for primetime / making it PR ready beyond the initial agent-driven first pass so I didn't want to open a PR just yet, but I thought I'd at least see if this was something you've considered or that you think would be a useful addition.
Cheers ✌️
Hi there, thanks for making this plugin! I've found it super useful so far. A few weeks ago while I was first getting started with it, I had the idea that in addition to tech spec and PRD modes, it could be useful to apply the same iterative adversarial model to code reviews. I actually used the plugin to first create a spec for this and then implement it here: main...bgentry:adversarial-spec:bg/code-review
It worked pretty well the times I've tried it so far! Getting the agents to argue about which code review issues are legit, worth fixing, and how to fix them has been a good way to quickly improve complicated PRs. I definitely haven't put a lot of time into getting it ready for primetime / making it PR ready beyond the initial agent-driven first pass so I didn't want to open a PR just yet, but I thought I'd at least see if this was something you've considered or that you think would be a useful addition.
Cheers ✌️