On Fri, 28 Aug 2026, Xuneng Zhou <[email protected]> wrote: > Interestingly, although the current SOTA model can do good things > without complicated prompts, it still seems useful for complicated > tasks. Here's a prompt that I plan to give Sol a try.
I'm running something very similar with Fable, also focusing on the prioritization / restartability as it eats up usage limits quickly. It already ran ~30% of the planned first-round checks, and reported 22 issues with reproducers so far. Running it with another model seems like a good idea, most likely while they will have common discoveries, some of the findings will be different.
