## Description > AI-agent disclosure: this PR was authored by an automated agent (Codex) and has not been line-by-line reviewed by a human before submission. Regex character ranges generate characters with `String.fromCodePoint`, but their unmapper uses `charCodeAt`, which returns only the first UTF-16 code unit. Valid astral characters are therefore rejected when shrinking without a generation context. Use `codePointAt` to invert the mapper correctly. Add regressions for literal and escaped Unicode ranges, checking that shrinking stays within the range and rejects invalid values. Validation: the stringMatching suite passes (8 tests); both new astral shrinking cases fail on the original implementation. Package typecheck, formatting, and diff checks pass. Changed-file lint reports no errors, with existing string-spread warnings. Patch changeset included. ## Checklist — _Don't delete this checklist and make sure you do the following before opening the PR_ - [ ] I have a full understanding of every line in this PR — whether the code was hand-written, AI-generated (our [AI Policy](https://github.com/dubzzz/fast-check/blob/main/AI_POLICY.md)), copied from external sources or produced by any other tool - [ ] I flagged the impact of my change (minor / patch / major) either by running `pnpm run bump` or by following the instructions from the changeset bot - [ ] I kept this PR focused on a single concern and did not bundle unrelated changes - [ ] I followed the [gitmoji](https://gitmoji.dev/) specification for the name of the PR, including the package scope (e.g. `🐛(vitest) Something...`) when the change targets a package other than `fast-check` - [ ] I added relevant tests and they would have failed without my PR (when applicable) <!-- PRs not checking all the boxes may take longer before being reviewed --> <!-- More about contributing at https://github.com/dubzzz/fast-check/blob/main/CONTRIBUTING.md -->