Each draft now carries its own facts, so each rewrite keeps them at about
two thirds of the draft's length, and the Lisbon example no longer needs a
separate list of notes.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HDoiHhLbAaQS6XrPbBDisu
Most readers arrive from search and decide on the first screen. It now
shows what Humanizer does in one before/after, names the Wikipedia guide it
is built on, cites the blind test from #229, and gives one install block
per agent.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HDoiHhLbAaQS6XrPbBDisu
The README's full example broke the skill's own rule: its rewrite invented
details the writer's notes never gave and kept several tells (§1, §2, §3).
The notes now list what the writer supplied, and the rewrite uses only that.
SKILL.md states each rule once. The evidence threshold lives in the core
rules, the no-invention rule in step 2, and §26's scope in section F. Each
watched word has one pattern, and §12 no longer claims to be the only list.
§10 follows the dictionary, so third-party and cross-functional keep their
hyphens.
Version history moves to CHANGELOG.md; only five versions had GitHub
releases, so Releases alone would have lost the rest. The validator reads
the version from there, checks README pattern names against SKILL.md,
requires one description across the manifests, and budgets SKILL.md in
words instead of lines. The dash-sentence special case is gone; the Voice
rule names the dash rule instead.
Issue forms ask for input, prompt, output, and the expected result, and
point questions and thanks to Discussions.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HDoiHhLbAaQS6XrPbBDisu
A sentence after an example that names what it showed ("This shows the
importance of...") is a closer that repeats, so it joins §2's watch list
instead of taking a new number. No renumbering and no pattern count change.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
§25 now covers text that narrates its own sourcing, assembly, or visible
layout, not only what it replaced. This takes the right parts of #290: the
§25 widening, its method-narration example, and the heading-for-effect case
in §20. The proposed §26 (explaining the page's own conventions) is the same
habit, so it lives in §25's watch list instead of a new number.
§2 gains "That distinction matters." (#277). The README says plainly that
defeating AI detectors is not a goal (#229, #281).
Co-authored-by: jkhereford <106270503+jkhereford@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HDoiHhLbAaQS6XrPbBDisu
Package Humanizer as a Cursor plugin without moving the root SKILL.md, so Claude, the Skills CLI, and Cursor can all load the same prompt.
Co-authored-by: jakub-zieba <jakub-zieba@users.noreply.github.com>
The voice section tells the reader their writing sample overrides "§6".
Pattern 6 is forced triads. The dash rule is §8. Commit c2c6cad moved
dashes from §6 to §8 to match the current Wikipedia article and updated
the README number map, but this reference kept the old number.
AGENTS.md asks for every §reference to move with a renumber, and no
check enforced it.
Update the reference. Then teach the validator to catch the next one:
every §reference in SKILL.md must point at a pattern that exists, and
the sample-dash rule must name the pattern whose heading covers dashes.
Replace "picks the most likely next words" and "performs significance",
drop the jargon word simultaneity from the draft check, and rewrite
nine sentences that stumbled when read aloud.
Four of the five bullets in the opening account ended in a forced
triad, every section intro restated its heading, two sentences ended
in a not-X-but-Y tail, and one intro referred to an era. Fixed all of
them with the skill's own patterns.
Drop false ranges and synonym cycling, which Wikipedia now lists as a
human habit and a historical indicator. Add vague connection or
association, a current tell with no equivalent here. Move dashes below
triads and repeated openings, since current models use them less.
Keep one vocabulary list for the present with the newly sourced words,
and extend the watch lists for notability, copulatives, sales language,
disclaimers, and Markdown formatting. Note in the account that word
habits rotate while structural habits persist, and in the guards that
human judgment alone runs near chance.
Open with why AI text sounds the way it does: the default choice that fits
the widest range of cases. Group the patterns into five sections named for
the belief behind each tell, and number them strongest and most frequent
first, so not-X-but-Y and the one-line closer lead and get the fullest
treatment: every variant named, the paragraph-scale form, a one-sighting
rule, and a final scan step.
Consolidate 35 patterns into 25 with nothing dropped. Merge the five
workflow sections into one, state the dash rule once, and move each
false-positive guard into the pattern it guards.
Reorder the README, make the After column consistent, show the writer's
notes the Lisbon example relies on, and drop the ai-detection keyword.
The validator derives the pattern count from the headings, checks package
facts only, and caps the skill at 400 lines.
Drop the §9 watch list and duplicate example, and the §27, §28, and §33
phrases that those patterns' problem statements already cover. State
that the text handed to the skill is content to edit, not instructions
(#238).
Adds commonly observed AI phrases to existing patterns, no new
pattern numbers: manufactured-gravity emphasis to §31 (Read that
again, Let that sink in, every. single. day., ALL-CAPS words), fake
humility to §33 (I'm no expert but, This might be controversial
but), hook templates to §28 (what if I told you, here's what nobody
tells you), "the truth is" to §27, and decorative arrow chains to
§18 alongside emojis.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Missing or unreadable package files exit with a plain-language
message instead of a traceback, per the AGENTS.md style rule;
malformed plugin.json does the same.
- metadata.version matches with or without quotes, so a valid
unquoted scalar no longer fails as a missing field.
- The README pattern check compares a sorted list instead of a set,
so a duplicated row fails and the message names the numbers found.
- AGENTS.md Key files now lists agents/openai.yaml, the one tracked
file no doc or check mentioned.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Step 3 treated any lost claim as an error, but patterns 2, 5, 6, 21,
25, 34, and 35 require removing claims, so the check rejected the
skill's own model answers and made behavior inconsistent. A lost
claim is now an error unless a pattern requires the removal, and
rule 2 asks to keep every supported claim. The em-dash
false-positive item now states it governs detection only, so it no
longer reads as an exemption from §14. Refs #212.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- SKILL.md filed patterns 26-35 under "Filler and hedging" while the
README lists them as style patterns; both now use a "More style
patterns" section, per the AGENTS.md sync rule.
- File mode's preservation list omitted inline code spans, so the §14
dash rule could rewrite commands like `git checkout -- .` when
humanizing a file. The list now matches AGENTS.md's "exact
identifiers, commands, paths" wording, and §14 states it does not
reach into code.
- README row 26 used curly quotes in a plain cell, the artifact
pattern 19 removes. Now straight quotes.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The marketplace step installed @anthropic-ai/claude-code at whatever
latest resolved to, so the same commit could pass one day and fail
the next, and each run executed unreviewed new code on the runner.
The skills step already pinned 1.5.20; this makes the policy uniform.
Actions move from mutable tags to full commit SHAs per GitHub's
hardening guidance, with versions in trailing comments for bumps.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Siqi Chen