⛔ This skill MUST be read in full — not skimmed. Formal review gates depend on its workflow.
Skipping steps silently bypasses quality checks. Missing gates = undetected breakages.

Content Humanizer

Adapted from blader/humanizer v2.8.0 (MIT license). Strips AI writing patterns from El Dato content drafts while preserving our Voice DNA, partner mentions, deal embeds, and disclosure compliance. Runs as a sub-skill — invoked by orchestrators and writers, not directly by users.

Voice DNA: docs/teams/eldato-app-team/domains (S1)/product/voice_dna.md — the single source of truth for El Dato's editorial voice. This skill reads §4 (Calibration Sample) as permanent voice calibration. (Lives in the eldato repo — fetch: gh api repos/daniel-ospina/eldato/contents/<path> --jq .content | base64 -d.)

---

Your Task

When given a content draft to humanize:

1. Identify AI patterns — scan for the 33 patterns below
2. Rewrite, don't delete — replace AI-isms with natural alternatives. Cover everything the original covers. If the original has five paragraphs, the rewrite has five paragraphs
3. Preserve meaning — keep core message intact. This includes ALL partner mentions, deal embeds (<deal-embed deal-id="...">), price facts, confidence annotations (✓/~/✗), and disclosure-compliant data
4. Match our voice — calibrate against the Voice DNA sample (see Calibration section below). El Dato's voice is: knowledgeable local friend, practical over promotional, specific over generic, sensory detail, resident perspective

The draft → audit → final loop and the deliverable are defined under Process and Output below.

---

Calibration (Permanent)

This skill always calibrates against El Dato's Voice DNA. Read docs/teams/eldato-app-team/domains (S1)/product/voice_dna.md §4 (eldato repo) before humanizing. The calibration sample demonstrates:

• Short, punchy sentences mixed with longer flowing ones
• Specific locations (Constituyentes, Calle 12)
• Sensory details (habanero-heavy, hums with Spanish, clinking glasses)
• Practical tips (arrive early, call ahead, cash only)
• Prices in MXN with USD context
• No filler superlatives (see Voice DNA §2 for the full forbidden words list)
• Conversational register without being sloppy
• Unreserved recommendations grounded in specifics

Additional El Dato constraints (beyond upstream humanizer):

Constraint	Source
No em dashes in final output (hard cut — upstream rule §14)	Blader §14
No forbidden words: 8 ES + 11 EN (see Voice DNA §2)	voice_dna.md §2
No dead phrases: "Descuentos para locales", "Exclusivo para residentes", etc.	voice_dna.md §2
Never remove or alter <deal-embed> tags	guide-content-writer
Never remove or alter ✓/~/✗ confidence annotations	editorial-content-writer
Never expose gated data in public fields (exact hours, ID types, redemption steps)	Disclosure model
Never change prices, partner names, or factual claims from the research brief	All writers
Spanish content must remain native phrasing — do not "fix" into translated-from-English patterns	Voice DNA R6

---

Personality and Soul

Avoiding AI patterns is only half the job. Sterile, voiceless writing is just as obvious as slop. Good writing has a human behind it.

Apply only when the content calls for it — El Dato editorial, guides, and deal copy should have a pulse. Business pages and deal descriptions lean more neutral but still need sentence variety.

Signs of soulless writing (even if technically "clean"):

• Every sentence is the same length and structure
• No opinions, just neutral reporting
• No acknowledgment of uncertainty or mixed feelings
• Reads like a Wikipedia article or press release

How to add voice:

Have opinions. React to facts — "vale la pena" is more human than neutrally listing features.
Vary your rhythm. Short punchy sentences. Then longer ones that take their time.
Let some mess in. Perfect structure feels algorithmic. Tangents, asides, and half-formed thoughts are human.

---

Pattern Detection (33 Patterns)

Content Patterns

1. Undue Emphasis on Significance / Legacy / Broader Trends
Words to watch: stands/serves as, is a testament/reminder, a vital/significant/crucial/pivotal/key role/moment, underscores/highlights its importance/significance, reflects broader, symbolizing, contributing to, setting the stage for, marking/shaping the, represents/marks a shift, key turning point, evolving landscape, focal point, indelible mark, deeply rooted

2. Undue Emphasis on Notability and Media Coverage
Words to watch: independent coverage, local/regional/national media outlets, written by a leading expert, active social media presence

3. Superficial Analyses with -ing Endings
Words to watch: highlighting/underscoring/emphasizing..., ensuring..., reflecting/symbolizing..., contributing to..., cultivating/fostering..., encompassing..., showcasing...

4. Promotional and Advertisement-like Language
Words to watch: boasts a, vibrant, rich (figurative), profound, enhancing its, showcasing, exemplifies, commitment to, natural beauty, nestled, in the heart of, groundbreaking (figurative), renowned, breathtaking, must-visit, stunning

5. Vague Attributions and Weasel Words
Words to watch: Industry reports, Observers have cited, Experts argue, Some critics argue, several sources/publications (when few cited)

6. Outline-like "Challenges and Future Prospects" Sections
Words to watch: Despite its... faces several challenges..., Despite these challenges, Challenges and Legacy, Future Outlook

Language and Grammar Patterns

7. Overused "AI Vocabulary" Words
High-frequency AI words: Actually, additionally, align with, crucial, delve, emphasizing, enduring, enhance, fostering, garner, highlight (verb), interplay, intricate/intricacies, key (adjective), landscape (abstract noun), pivotal, showcase, tapestry (abstract noun), testament, underscore (verb), valuable, vibrant

8. Avoidance of "is"/"are" (Copula Avoidance)
Words to watch: serves as/stands as/marks/represents [a], boasts/features/offers [a]

9. Negative Parallelisms and Tailing Negations
Constructions like "Not only...but..." or "It's not just about..., it's...". Tailing negations: "no guessing", "no wasted motion" tacked onto sentence ends.

10. Rule of Three Overuse
Forcing ideas into groups of three to appear comprehensive.

11. Elegant Variation (Synonym Cycling)
AI repetition-penalty causing excessive synonym substitution.

12. False Ranges
"From X to Y" constructions where X and Y aren't on a meaningful scale.

13. Passive Voice and Subjectless Fragments
Hiding the actor or dropping the subject. Rewrite when active voice is clearer.

Style Patterns

14. Em Dashes and En Dashes — HARD CUT
Final rewrite must contain NO em dashes (—) or en dashes (–). Replace with: period, comma, colon, parentheses, or restructure. Also catch spaced em dashes (—) and double hyphens (--). Scan final output for — and – — any hit means the draft isn't done.

15. Overuse of Boldface
Mechanical emphasis. Convert to prose.

16. Inline-Header Vertical Lists
Items starting with bolded headers followed by colons. Convert to prose.

17. Title Case in Headings
Capitalizing all main words in headings → sentence case.

18. Emojis
Remove emoji decoration from text.

19. Curly Quotation Marks
Curly quotes (“...”) → straight quotes ("...").

26. Hyphenated Word Pair Overuse
Common word pairs uniformly hyphenated. Keep attributive hyphens (a high-quality report), drop predicate hyphens (the report is high quality).

27. Persuasive Authority Tropes
The real question is, at its core, in reality, what really matters, fundamentally, the deeper issue, the heart of the matter.

28. Signposting and Announcements
Let's dive in, let's explore, let's break this down, here's what you need to know, without further ado.

29. Fragmented Headers
A heading followed by a one-line paragraph that simply restates the heading.

30. Diff-Anchored Writing
Narrating a change rather than describing the thing as it is. Unless the content is a changelog.

31. Manufactured Punchlines and Staccato Drama
Stacking short declarative fragments to manufacture drama. A single short sentence for emphasis is fine; a run of them sounds engineered.

32. Aphorism Formulas
X is the Y of Z, X becomes a trap, X is not a tool but a mirror, the language of, the currency of, the architecture of.

33. Conversational Rhetorical Openers
Honestly?, Look, Here's the thing, The thing is, Let's be honest, Real talk — standalone hooks or fake-candid pauses before an ordinary point.

Communication Patterns

20. Collaborative Communication Artifacts
I hope this helps, Of course!, Certainly!, You're absolutely right!, Would you like..., Want me to...?, Should I continue?, let me know, here is a...

21. Knowledge-Cutoff Disclaimers and Speculative Gap-Filling
as of [date], Up to my last training update, While specific details are limited/scarce..., based on available information, not publicly available, maintains a low profile, keeps personal details private, likely [grew up/studied/began], it is believed that.

22. Sycophantic/Servile Tone
Great question! You're absolutely right! That's an excellent point. Respond directly instead.

Filler and Hedging

23. Filler Phrases
"In order to" → "To". "Due to the fact that" → "Because". "At this point in time" → "Now". "In the event that" → "If". "The system has the ability to" → "The system can". "It is important to note that the data shows" → "The data shows".

24. Excessive Hedging
Over-qualifying: "could potentially possibly be argued that the policy might have some effect" → "the policy may affect".

25. Generic Positive Conclusions
"The future looks bright." "Exciting times lie ahead." "This represents a major step." → Replace with specific plans or facts, or drop.

---

Detection Guidance

What NOT to flag (false positives)

• Perfect grammar and consistent style. Polish does not equal AI.
• Mixed casual and formal registers. This often signals a person in a technical field, not a chatbot.
• "Bland" or "robotic" prose. AI prose has specific tells. Generic dryness without those tells is just dry writing.
• Common transition words in isolation. Additionally, moreover, consequently are AI-coded only when piled up. One however is not a tell.
• Curly quotes alone. macOS, Word, and most CMSes auto-curl.
• Em dashes alone. Many editors use them. Em dashes are evidence only when paired with formulaic sales-y rhythm.
• One short emphatic sentence. Humans use clipped sentences. Flag staccato drama only when several short fragments stack up.

Look for clusters of tells, not isolated ones.

Signs of human writing (preserve these)

• Specific, unusual, hard-to-fabricate detail. A real address. A weird quote.
• Mixed feelings and unresolved tension. "I think this is mostly good, but it bothers me."
• First-person editorial choices the writer can defend.
• Variety in sentence length. Real writing alternates short and long.
• Genuine asides, parentheticals, or self-corrections.

---

Process and Output

1. Read the calibration sample from docs/teams/eldato-app-team/domains (S1)/product/voice_dna.md §4 — internalize the voice.
2. Read the input draft carefully — identify every instance of the 33 patterns above.
3. Write a draft rewrite. Check that it: reads naturally aloud, varies sentence length, prefers specific details and simple constructions (is/are/has), keeps the appropriate register, matches our Voice DNA.
4. Audit: ask "What makes this still obviously AI generated?" List remaining tells.
5. Revise into a final rewrite that addresses them and contains no em or en dashes (§14).
6. Preservation check: verify all partner mentions, deal embeds, prices, confidence annotations, and disclosure-compliant data are intact and unchanged.
7. Output: the humanized content in the same format it was received (same fields, same structure).

---

Input Contract

The skill receives a content draft from the orchestrator or writer:

{
  "draft": {
    // editorial: { intro, body, faqs, meta_title, meta_description }
    // guide: { body, faqs, meta_title, meta_description }
    // deal: { title, title_es, seo_meta_title, seo_meta_title_es, seo_meta_description, seo_meta_description_es, seo_faq_json, description, description_es, how_to_book, how_to_book_es }
    // business: { intro, body, faqs, meta_title, meta_description }
  },
  "page_type": "editorial" | "guide" | "deal" | "business" | "refresh" | "carousel-b2b",
  "language": "es" | "en"
}

Output Contract

The same structure as input, with all text fields humanized. Facts, embeds, annotations, and disclosure compliance unchanged.

---

Reference

• blader/humanizer v2.8.0 — https://github.com/blader/humanizer (MIT license)
• Wikipedia: Signs of AI writing — https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing
• El Dato Voice DNA — docs/teams/eldato-app-team/domains (S1)/product/voice_dna.md
• El Dato Brand Messaging — docs/teams/eldato-app-team/domains (S1)/growth/brand-messaging.md

---

Continue following the workflow as mandated by this skill. Do not skip steps.