@OpenAIDevs@thsottiaux when the models run long tasks, such as benchmarks or downloading something or stuff like that, would be nice interrupting the interaction and reactivating it when the long task is completed instead of printing progress and burning budget. Thx
@Bugcrowd I opened 6 or 7 bugs in these days, some are p1 and p2. All marked as "Not Applicable". Triagers saying there are not enough proofs, when I have uploaded full .zip files. Are you kidding? Worst bb platform ever!
@0xConda I am getting a tremendous amount of duplicates, idk how you guys are able to submit non-duplicate stuff. The last one critical 9.0 was submitted 26 mimutes before mine...
Just found a CRITICAL 9.0 on a famous program and guess what? Marked as duplicate because someone submitted the same vulnerability 26 minutes before. I have a number of duplicates that you can not imahine, something like 40 in 2 months.