Rewrite Humanizer in Plain Language - #223
Merged
Merged
Conversation
blader
marked this pull request as ready for review
August 17, 2026 10:04
ztemerbekov
added a commit
to ztemerbekov/a1-marketing-skills
that referenced
this pull request
Aug 17, 2026
## What changed - replace the compact A1 runtime adaptation with the upstream Humanizer 2.11.0 `SKILL.md` - preserve the original 35 patterns, examples, voice calibration, false-positive guard, and draft → audit → final workflow verbatim - remove A1's former final-only default, marketing-strategy boundary, and custom language-independent pattern catalog - remove upstream file mode so the skill never reads or overwrites project files or Marketing Context - keep only required A1 integration differences: skill name/frontmatter schema, language contract, icons and client metadata, source boundary, support footer, and attribution - use upstream `author.name` (`blader`) for `metadata.author`; keep MIT terms and Siqi Chen's full attribution in `references/license-and-attribution.md`, not in runtime frontmatter - align the English and Russian documentation and evaluation criteria with the upstream behavior ## Why A1 Humanize was intended to port Humanizer while removing file access. The previous implementation instead rewrote and narrowed the upstream runtime, including locally chosen Russian vocabulary and a different output contract. This change corrects that divergence and syncs the skill with the large plain-language update in blader/humanizer#223. ## User impact Users get Humanizer 2.11.0 behavior directly: the original pattern definitions and examples, original voice and false-positive handling, and the original pasted-text output of draft, short audit, and final rewrite. Explicit final-only requests and embedded use still return only the rewrite. ## Verification - the SHA-1 of the runtime block from `Match the writer's voice` through `How to return the result` matches upstream exactly: `6b55e577e9d8f9b601812e03bed6cc311355b1c9` - `bash -n scripts/validate-skills.sh` - `./scripts/validate-skills.sh` (6 skills, 97 eval cases, 0 failures) - `git diff --check`
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What changed
SKILL.mdin Plain Language and shortened it by about 750 words.AGENTS.mdand made the package validator check it.v2.11.0.What stayed the same
SKILL.mdremains the only prompt authority.Why
Humanizer should model the clear writing it asks other agents to produce. The new wording leads with the main point, uses common words and active voice, keeps sentences short, and explains necessary technical ideas.
Checks
python3 scripts/validate-package.pypython3 .../skill-creator/scripts/quick_validate.py .npx --yes skills add . --listclaude plugin validate .maingit diff --check