What proof should a coding agent show before another retry?
by•
One pattern we keep seeing: teams set a max budget or max iterations, but the expensive part usually starts earlier, when the agent keeps retrying without new evidence.
A better stop rule seems to be: before another retry, the agent should show what changed, what verifier passed, and what would make the next attempt stop.
I am curious what your minimum receipt would be here. If you run Claude Code, Codex, or another coding agent, what proof would make you comfortable letting it continue instead of stopping the loop?
6 views
Replies
The simplest version I have seen work is: before any retry, show one sentence for what changed, one verifier result, and one reason the next attempt is still worth budget. Without that, a retry usually just feels like hope with tokens attached.