If Developer Assist is outputing code that exceeds its limits, it truncates. I suggest before printing to check whether the output has exceeded its limit. If so, use a larger model, chunk, etc.
Right now, you’re forced to ask Developer Assist to address smaller chunks.
Outputing partial code provides no use to the developer, so ensuring complete context to what the assist is trying to show the developer could be invaluable. Have the output dynamically adjust model usage based on token usage requirements (within reason).
@MSL0727 - thank you for sending along this valuable feedback! We just recently implemented improvements that have increased token limits across the board so this should not happen again. This is currently going through the release process so may not yet be in place on your environment but will be soon. Please let us know if you’re still experiencing friction here after the updates.
@MSL0727 - following up as this been released for some time, are you able to confirm that you have no longer experienced this behavior and can consider this resolved?
