@mattpocockuk , idea I had that seems to be working pretty well. When using /grill-with-docs, in my prompt to start the grilling I add "Should we use ...."
I'm noticing that is very clever about if it should or not as it's so far really good at estimating if it has enough info.
@scaling01 OpenAI is not allowed to run claude for their benchmarks. Though they really should burry the hatchet... I'm not saying become friends. But stuff like this is annoying.
So many announcements and updates from the camp of @OpenAI. But no GPT 5.6.
Are they waiting for a result of the fable - america situation before releasing it. Pushing other updates in the meantime?
@thsottiaux , 3rd time I lost all chats of a particular project on codex π , sadly don't know what caused it. Frustrating experience.
When it does happen it seems to be for multiple projects, but not all? Hope you can get it fixed π
Combined @mattpocockuk's /grill-with-docs and /to-issues today, and it was a really pleasant experience.
I had a clear sense of what it was going to do beforehand, during, and afterward I felt like I fully understood what it had created. Really nice workflow.
Hey @BNPPARIBASOPEN , I can't make a payment to @cursor_ai because you think the site is fraudulent.
Cursor is not a fraudulent company!
Could you look into it and resolve it please? I've had multiple interactions with your support and it's not resolved yet.