A long failing trace is useful evidence, but it is not always a useful regression test. Most of the events may have nothing to do with the bug. CaseCrop removes events, preserves their declared prerequisites and replays the remaining sequence. It keeps a candidate only when the original failure signature comes back. A different crash is not a successful reduction. The browser workbench runs the Python library locally through Pyodide. The bundled cache example reduces 36 events to six. You can inspect the attempted reductions and export the evidence; the CLI can also generate an executable regression case. It is free and MIT licensed, with no runtime dependencies. To use your own application, provide a replay function that resets state each time. The reported minimality is closure 1-minimality, not a global minimum or a proof of root cause. If you debug stateful workflows, what would make connecting your replay harness easier? Try it: https://casecrop.web.app Source: https://github.com/shi1720/casecrop Built by Shivam Gupta. More work and contact: https://shivamgupta.web.app/ LinkedIn: https://www.linkedin.com/in/shiv...

CaseCrop

Turn a long failing trace into a smaller regression case

Turn a long failing trace into a smaller regression case