Rewind a failing test to hunt down a software bug
Antithesis let Claude use its rewind tool to replay a failing test and probe a bug in HashiCorp software; the company shows a replay but reports no final fix.
Done with Claude
What they did
Antithesis, a software testing company, let agents use its tool that can replay and rewind a test run exactly. In their demo, Claude takes a bug Antithesis found in HashiCorp's Raft, software that keeps servers in agreement. It rewinds to just before the failure and tries ideas in separate copies of the run. Then it reruns the changed code to check the fix.
How it went
The company shows a replay of Claude hunting the bug. They say a person no longer needs to run the tool and pass findings to the agent. No fix results are given.
Worth knowing
You need an Antithesis account (Release 64) and existing Antithesis tests that are already failing. It is a paid product, and the post is its launch announcement.
Try it yourself with Claude
Using the antithesis-debug skill, take the failing property [property name] from Antithesis run [run ID]. Rewind to just before the assertion failed. Run diagnostic commands in branched timelines to find the root cause, and explain the sequence of events. Propose a fix and show me the diff before committing. Then launch a new Antithesis run to check that the fix holds.
Swap the highlighted [parts] for your own details.
Source: Hacker News · David Koenitzer · Oct 15, 2026
Written with AI assistance from the original post and checked against our editorial guidelines. Results are as reported by the source, not independently verified.
