Skip to content

docs: add real-world before/after reference example#69

Open
Kholtien wants to merge 2 commits intoblader:mainfrom
Kholtien:add-reference-examples
Open

docs: add real-world before/after reference example#69
Kholtien wants to merge 2 commits intoblader:mainfrom
Kholtien:add-reference-examples

Conversation

@Kholtien
Copy link
Copy Markdown

Summary

  • Adds references/ directory with a before/after example from a real published AI-generated article (KRS Channel astrology article from Breeze Astrology)
  • Scanner scored the original 12.6/100w (MODERATE, 53 hits) and the humanised version 0.0/100w (clean, 0 hits)
  • The original was 875 words; the humanised version is 323 words with actual opinions and specifics
  • Updates SKILL.md to document the reference examples

Depends on #68 — this PR builds on the pattern detection script commit.

Test plan

  • Verify references/krs_astrology_article_before.txt includes source attribution
  • Run python3 scripts/detect_patterns.py references/krs_astrology_article_before.txt and confirm score of 12.6/100w
  • Run python3 scripts/detect_patterns.py references/krs_astrology_article_after.txt and confirm score of 0.0

🤖 Generated with Claude Code

Kholtien and others added 2 commits March 23, 2026 12:57
Adds scripts/detect_patterns.py — a stdlib-only Python script that scans
text for measurable AI writing patterns and produces a scored report.
Detects 15 of the 25 patterns programmatically (AI vocabulary, em dashes,
boldface, emojis, filler phrases, hedging, and more). Outputs a normalised
score per 100 words for objective before/after comparison.

Updates SKILL.md (v2.4.0) to integrate the scanner into the humanisation
process at three points: pre-scan, post-scan, and final scan. Updates
README.md with script documentation and usage examples.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds references/ directory with a before/after example from a real
published AI-generated article (KRS Channel astrology article from
Breeze Astrology). Scanner scored it 12.6/100w before and 0.0 after
humanisation. Updates SKILL.md to document the reference examples.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Copy link
Copy Markdown

@Aniol0012 Aniol0012 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Interesting, but I would keep the SKILL.md as standalone as possible.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants