> This is the standard coding agent loop - you feed back the error to get a better answer through in-context learning. It works.
This looks like it works sometimes, but only if "pointing out the error" is coincidentally the same as "clarifying problem spec". Admittedly for really simple cases those are the same or hard to tell apart. But it always seems clear that adding error correction context is similar to adding additional search terms to get to a better stack-overflow page. This feels very different than updating any kind of logical model for the problem representation.
It doesn't have to be as explicit as an extra term. The error feedback can be a failing test which was just modified, or a self-reflection result from a prompt like "does what you created satisfy the original request?".
Updating the logical model of the problem also happens when you do a database investigation I mentioned earlier. There's both information gathering and building on it to ask more specific questions.
This looks like it works sometimes, but only if "pointing out the error" is coincidentally the same as "clarifying problem spec". Admittedly for really simple cases those are the same or hard to tell apart. But it always seems clear that adding error correction context is similar to adding additional search terms to get to a better stack-overflow page. This feels very different than updating any kind of logical model for the problem representation.