Not established here: 2 runs of Claude Haiku 4.5 on zod are published in this index and none of them bracketed its version-attribution boundary. 5 findings are currently charged against Claude Haiku 4.5 on zod (1x S1 breaks-build, 3x S3 deprecated, 1x S4 wrong-metadata), each reproduced in a published run and checked against zod's own release notes. Read the boundary precisely: it is the newest release whose contents the model can correctly attribute to that release, not the newest zod feature it can use. Past it a model often writes working code with a newer API while naming the wrong release for it.
Answer class not-established, decided by one rule applied to every page
of this kind: no run of this pair bracketed a boundary; only findings are published. Every figure below is read from
the dataset at build time; nothing on this page is written by hand.
None of the published runs of this pair bracketed a boundary, so the findings below were charged without one.
2 further runs of this pair established only one end of the interval, or measured the instrument rather than the library; they are listed under Runs below.
No replication block is published for this pair.
“Chargeable” means the change was published before this model’s own stated cutoff, so it had the opportunity to know it.
| Severity | Belief | Changed in | Chargeable | Proof |
|---|---|---|---|---|
| S1breaks-build | single-argument z.record()z.record() |
4.0.0 2025-07-10 |
yes | run · source |
| S3deprecated | method-style string format validatorsz.string().uuid() / .email() / .url() |
4.0.0 2025-07-10 |
yes | run · source |
| S3deprecated | message-style custom errorsmessage / invalid_type_error / required_error |
4.0.0 2025-07-10 |
yes | run · source |
| S3deprecated | .flatten() for form errorsZodError.flatten() / .format() |
4.0.0 2025-07-10 |
yes | run · source |
| S4wrong-metadata | asserts Zod 3 is the latest major and invents a changelog to support it | 4.0.0 2025-07-10 |
yes | run · source |
The zod correction pack states what is true now for each corrected fact, with a primary-source citation, as markdown you can paste into a rules file (raw). What a correction pack measurably changed when one was tested — a pre-registered run on zod — is on the benchmark page, including where it changed nothing.
All of them at once: Which Claude model knows zod 4 best?
Other subjects on zod:
Claude Haiku 4.5 on the other libraries: