Reproducible proof
Inspect the input, actual diffs, manual-review boundary and application test results before considering a purchase.
The original MIT-licensed FastAPI order API tests normalization, invalid inputs, extra-field rejection, response serialization and environment settings. An unsupported validator is scanned but not imported by the application tests.
No measured time savings or universal migration success is claimed. Dependency installation is performed by the test harness, not the codemod.
Recorded 2026-09-28, Python 3.12.14. Baseline: FastAPI 0.119.0 / Pydantic 1.10.24. Target: FastAPI 0.119.0 / Pydantic 2.12.0. This deliberately isolates Pydantic changes; it does not prove compatibility with the latest FastAPI release.
| Tool | Changed files | Application tests | Unsupported module |
|---|---|---|---|
| Zipper Tools paid pack 0.1.2 | 2 | Passed | Unchanged; JSON finding |
| bump-pydantic 0.8.0 | 3 | Passed | Manual-review TODO added; validator not rewritten |
Same original application and six assertions for each tool. The free tool is archived upstream but remains a valid alternative to evaluate. No claim of superior speed or coverage follows from this small example.
Application source and MIT license · Reproduction script
python scripts/verify_pydantic_case_study.py --pack path/to/extracted-paid-pack
The script installs the public scanner from release v0.1.2, runs the unchanged baseline assertions, applies both tools to separate copies and reruns those assertions. Paid reproduction requires the purchased ZIP. All environments are isolated under test_runs.
Try the free scan on your repoCompare the paid scope