humanizer 1.2.1
Documentation-only patch. This release clarifies exactly which concepts from
watermarks-remover Humanizer adopts without changing the humanization or
text-hygiene workflow.
Changed
- Added a feature adoption matrix to the README.
- Documented prompt-level invisible Unicode cleanup as the adapted feature.
- Documented statistical token-pattern disruption as partial, incidental, and
unverifiable. - Confirmed that C2PA, EXIF, XMP, PDF and document metadata, media processing,
and directory or website audits are not included. - Bumped the documented skill version to 1.2.1.
Validation
- Skill validator passed.
- Evaluation JSON parsed successfully.
- Git whitespace checks passed.
Full changelog: v1.2.0...v1.2.1