Retrace records every LLM and tool call in an agent run, lets you replay it step by step like a video, and lets you fork any step to change the input and see how the run plays out differently. I built it because debugging an agent that failed in prod usually means re-running the whole thing and guessing, and I wanted to reproduce the exact run and change one step instead. Docs Playground
When you replay or fork a run in Retrace, the steps before the fork come from the recording, but everything after runs live against the model. So two runs of the same input rarely match exactly, even when nothing actually broke.
That makes the useful question harder than it sounds: when a replay diverges, is it a real regression from your change, or just provider non-determinism? Retrace currently shows a first-divergence diff and a verdict of improved, regressed, or unchanged, but I would like to hear how others handle it. What tolerance do you use in practice, and would you rather see a strict step-by-step diff or a semantic comparison of each step?
Record, replay, fork & share AI agent executions. See every LLM call, tool invocation, and error your agent makes, then debug and iterate in seconds. Free for 1,000 traces/mo.