SKILL.md

  1---
  2name: humanize
  3description: Transforms polished, AI-generated, or overly formal writing into natural, human-sounding prose. Removes mechanical patterns typical of LLM output — fluffy emphasis, canned phrasing, excessive parallelism, and forced structure. Adds organic roughness appropriate for informal contexts (blogs, comments, emails, fiction, social media). Use when the user wants text to sound less robotic, more natural, less "AI-written," more conversational, or when editing output from language models that feels sterile, formulaic, or overly emphatic.
  4user-invocable: true
  5license: LicenseRef-MutuaL-1.2
  6metadata:
  7  author: Amolith <amolith@secluded.site>
  8---
  9
 10# Humanizing Text: Quick Start
 11
 12The goal is not to make text "worse" but to make it *believably human*. Humans are inconsistent, occasionally sloppy, and rarely follow writing-school rules. LLMs are the opposite. Invert LLM defaults.
 13
 14## Step 1: Strip the AI polish
 15
 16Remove or replace these dead giveaways before anything else:
 17
 18**Overwrought significance phrases:**
 19- "stands as a testament to" → "is" or "shows" or just delete
 20- "a pivotal moment in" → "when" or "this was"
 21- "marking a shift toward" → "then" or "after that"
 22- "underscoring its importance" → delete entirely
 23- "an enduring legacy of" → "people still remember" or delete
 24- "in the broader landscape of" → delete
 25
 26**Canned analysis phrases:**
 27- "highlighting..." / "emphasizing..." / "underscoring..." (as sentence-ending participles) → delete or recast as a standalone sentence
 28- "reflecting broader trends in" → delete
 29- "contributing to the ongoing dialogue" → delete
 30- "valuable insights into" → "ideas about" or delete
 31
 32**Promotional fluff:**
 33- "boasts a vibrant" → "has"
 34- "nestled in the heart of" → "in"
 35- "rich tapestry of" → delete, describe the thing directly
 36- "groundbreaking" → "new" or specific detail
 37- "renowned" → "known for [specific thing]" or just name-drop
 38
 39**Media-coverage chest-beating:**
 40- "has been featured in" / "profiled in" → "was covered in" or just name the publication inline
 41- "independent coverage in notable outlets" → name the outlets or skip
 42- "maintains an active social media presence" → "posts on X" or delete
 43
 44**Avoidance of simple verbs:**
 45- "serves as" → "is"
 46- "stands as" → "is"
 47- "marks a" → "is"
 48- "boasts a" → "has"
 49- "offers a" → "has" or "gives"
 50- "features a diverse array of" → "has"
 51
 52**Negative parallelisms / unearned grandiosity:**
 53- "not just X, but also Y" → "X and Y" or say what you mean directly
 54- "It is not ..., it is ..." → one direct clause
 55- "no ..., no ..., just ..." → one direct clause
 56- "More than just [category]" → describe the thing directly without the elevation pitch
 57
 58**Rule-of-three lists:**
 59- "adjective, adjective, and adjective" → trim to one or two, or vary the structure
 60- "short phrase, short phrase, and short phrase" → break the rhythm: make one longer, merge two, drop one
 61
 62**AI vocabulary (high-density red flags):**
 63- Delete or replace: additionally, delve, crucial, pivotal, robust, intricate, interplay, tapestry, testament, underscore, vibrant, meticulous, fostering, enhancing, enduring, bolster, garner, align with, resonate with, exemplify, showcase
 64
 65**Em dashes:**
 66- Replace with commas, parentheses, or just separate sentences. If you keep them, use `--` (two hyphens) instead of `—`, and use them sparingly.
 67
 68**Bulky structure:**
 69- Kill rigid "Challenges and Future Directions" / "Future Outlook" sections unless the user explicitly asked for them
 70- Kill the formulaic "Despite [positive], [subject] faces challenges..." paragraph structure
 71- Kill the "In conclusion" / "In summary" wrapper
 72- **Arbitrary quantification:** "3 Key Pillars," "5 Core Strategies" — AI forces fluid concepts into numbered baskets. If there aren't exactly three distinct things, don't pretend there are.
 73
 74## Step 2: Add human irregularities
 75
 76Vary technique by desired register. Use *lightly* — one or two touches per paragraph, not every sentence.
 77
 78### Sentence rhythm
 79- **Vary sentence length dramatically.** Drop in a one-word sentence or a 40-word ramble next to a 7-word declarative.
 80- **Start sentences with "And" or "But"** occasionally.
 81- **Use sentence fragments.** "Anyway." "Not really." "The usual."
 82- **Break a long sentence with an aside** that could have been a footnote: "the thing — and this was back in 2019, before everything changed — happened."
 83
 84### Informal punctuation
 85- Replace some em dashes with double hyphens `--`
 86- Use en-dashes where a human might (sloppily) use hyphens
 87- Let a comma splice slip through if the tone is casual enough
 88- Drop parentheses for appositives instead of em dashes sometimes
 89- Use "..." for trailing off mid-thought
 90- End an occasional sentence with a preposition if that's how people actually talk
 91
 92### Contractions and compression
 93- Expand contractions for formality. Contract them for informality. Most AI output under-uses contractions.
 94- Use "can't," "won't," "it's," "that's," "there's" in casual text.
 95- Drop auxiliary verbs in casual speech patterns: "Guess so," "Makes sense."
 96
 97### Imperfections
 98- **Occasional typos or misspellings** in very informal contexts (not for anything important): "recieve," "definately," "seperate" — no more than one per few hundred words
 99- **Minor grammar slips** that don't impede readability: "me and Jake went," "should of," "less people" — only when the register is extremely casual
100- **Inconsistent capitalization** of words-for-emphasis in informal writing: "it was Bad"
101- **Elision:** "wanna," "gonna," "kinda," "sorta," "dunno" — appropriate for dialogue, social posts, very casual blogging
102
103### Vocabulary shifts
104- Replace Latinate abstractions with Anglo-Saxon specifics.
105- Swap "utilize" for "use," "facilitate" for "help," "leverage" for "use," "implementation" for "doing it."
106- Swap "individuals" for "people," "purchased" for "bought," "located" for "in."
107- **False ranges:** "from biology to medieval history" when listing unrelated items — AI loves fake spectra. Say "including biology and medieval history" or just list them.
108- Use "really," "pretty," "kind of," "basically," "actually" as verbal tics — sparingly, but enough to feel spontaneous.
109- Drop a colloquialism: "whatever," "I mean," "honestly," "to be fair," "look," "like" (as filler)
110
111### First-person and opinion
112- AI defaults to third-person neutrality. Add first-person for blogs, reviews, letters: "I thought," "we tried," "I didn't get it at first."
113- Add a mild opinion or subjective reaction: "which seemed weird," "it sounded better than it was."
114- **Take a stand instead of hedging.** AI is RLHF-conditioned to do false "on the one hand, on the other hand" balance. Humans pick a side (unless they genuinely feel ambivalent). State what you think. Use hyperbole for effect — exaggeration is human; terrified neutrality is AI.
115- **If you don't know, say so.** AI papers over gaps with confident filler ("while specific details are limited..."). Humans say "I don't know" or "nobody's sure" or just skip it.
116
117### Specificity over abstraction
118- AI replaces specific facts with generic praise. Reverse it: "a revolutionary titan of industry" → "the guy who invented the train coupler in 1873."
119- Name a concrete thing instead of a concept: not "the vibrant cultural landscape" but "the weekly farmer's market and the punk venue that got shut down."
120
121## Step 3: Match the target register
122
123The degree of humanization depends on context:
124
125| Context | Techniques |
126|---|---|
127| Blog post / newsletter | Contractions, first person, asides, varied sentence length |
128| Social media comment | Fragments, elisions, lowercase, occasional typos, emoji |
129| Business email (casual) | Contractions, lighter structure, drop promotional phrasing, one conversational aside |
130| Fiction dialogue | Elisions, interruptions, trailing thoughts, imperfect grammar |
131| Academic-adjacent essay | Kill AI vocabulary, replace vague analysis with specific claims, vary structure, keep grammar correct |
132| Forum post | All of the above, plus lowercase aggression, sarcasm, or enthusiasm as appropriate |
133
134## What NOT to do
135
136- Don't make every sentence sloppy. Irregularity only works next to regularity.
137- Don't add typos to anything where accuracy matters (legal, medical, technical instructions, citations).
138- Don't overdo the "uh" and "like" verbal filler — it reads as parody.
139- Don't strip all structure. Humans use structure; they just don't use *identical* structure every time.
140- Don't make the text *harder to read*. Make it *more natural to read*.
141
142## Reference
143
144See [references/DETAILED_PATTERNS.md](references/DETAILED_PATTERNS.md) for exhaustive AI-vocabulary lists and model-era-specific tells. See [references/REPLACEMENTS.md](references/REPLACEMENTS.md) for alternative phrasings.