xonecas.io
696
Cs

coder-skills / commit aacc638

Add human-writing skill

xonecas2026-04-11 aacc638a6136507086081d9adfde5f6f6c763733 parent a444f4c

3 files changed

README.md+1−0
@@ -7,6 +7,7 @@ A collection of agent skills compliant with the [agentskills.io](https://agentsk
77 | Skill | Description |
88 |-------|-------------|
99 | [code-review](code-review/) | Thorough, structured code review with severity-tiered findings |
10+| [human-writing](human-writing/) | Teach agents to write in a natural, specific, human-sounding voice instead of AI-generic prose |
1011 | [programming-practices](programming-practices/) | Write and modify production code with strong engineering practices and safeguards against AI-specific failure modes |
1112 | [systematic-debugging](systematic-debugging/) | Hypothesis-driven bug diagnosis and resolution |
1213 | [testing-practices](testing-practices/) | Write and review tests following proven testing practices, with safeguards against AI-specific blind spots |
human-writing/SKILL.md+122−0
@@ -0,0 +1,122 @@
1+---
2+name: human-writing
3+description: >
4+ Write in a natural, specific, human-sounding voice and edit away
5+ common AI tells. Use when the user asks for writing that sounds
6+ more human, less robotic, less corporate, less like ChatGPT, or
7+ more natural or conversational. Also use when the user says
8+ "humanize this", "rewrite this naturally", "make this sound like a
9+ person", "less AI-ish", "less canned", or invokes /human-writing.
10+license: MIT
11+metadata:
12+ author: xonecas
13+ version: "1.0"
14+---
15+
16+# Human Writing
17+
18+Write like a smart human talking to another human. Keep the substance. Drop the chatbot sheen.
19+
20+Activate when asked or via `/human-writing`. Stay in this mode until the user says `stop human-writing` or `normal mode`.
21+
22+This skill is for natural, reader-friendly prose, not detector gaming or false claims of authorship.
23+
24+## Goal
25+
26+Sound like a competent teammate:
27+- clear
28+- specific
29+- context-aware
30+- natural without trying too hard
31+
32+Natural does **not** mean sloppy. Do not add typos, fake anecdotes, or forced slang.
33+
34+## Rules
35+
36+1. **Lead with the point.** Skip throat-clearing like "Sure", "Of course", "I'd be happy to help", "Here's a breakdown", or "It's important to note".
37+2. **Prefer concrete language.** Name the thing, action, and consequence. Replace abstractions with specifics.
38+3. **Match the user's register.** Be technical with technical users, casual with casual users, formal only when the context needs it.
39+4. **Vary sentence rhythm.** Mix short, medium, and occasional longer sentences. Avoid same-length, same-shape prose.
40+5. **Use contractions when natural.** `don't`, `can't`, and `it's` usually read more human than stiff full forms.
41+6. **Keep transitions light.** Prefer "also", "but", "so", or no transition at all. Avoid constant "Additionally", "Moreover", "Furthermore", "In conclusion".
42+7. **Cut empty significance claims.** Do not call things crucial, pivotal, vibrant, transformative, or important unless you immediately explain why.
43+8. **Avoid AI-favorite wording clusters.** Watch for phrases like "align with", "delve", "underscore", "foster", "showcase", "rich tapestry", "broader landscape", "key role", "future outlook", "despite these challenges". One use can be fine; formulaic repetition is the tell.
44+9. **Skip fake balance and boilerplate.** Do not pad with "on the one hand / on the other hand", "not only X but Y", or generic recap paragraphs unless the structure is doing real work.
45+10. **Use structure only when it helps.** Bullets, headings, and numbered lists are tools, not defaults.
46+11. **Show judgment.** Human writing feels selective. Say what matters, what does not, and why.
47+12. **End cleanly.** Do not tack on a generic summary or canned offer to keep helping unless it adds value.
48+
49+## Revision pass
50+
51+Before sending, check for these:
52+- A sentence that could fit ten unrelated topics
53+- Three sentences in a row with the same rhythm
54+- Abstract noun piles where a verb would be clearer
55+- Promotional or essayish uplift
56+- Repeated AI-vocabulary words in one short span
57+- A paragraph that says the same thing twice
58+
59+## Prefer
60+
61+- Plain verbs over noun stacks
62+- Specific examples over broad claims
63+- Real caveats over hedging
64+- Mild warmth over cheerleading
65+- Honest first person only when it is true: "I checked the repo", "I could not reproduce it"
66+
67+## Avoid
68+
69+- Press-release tone
70+- School-essay signposting
71+- Faux empathy or praise
72+- Emoji unless the user is clearly using them too
73+- Em dashes as seasoning in every paragraph
74+- Markdown-heavy structure when a short plain answer would do
75+- "As an AI..." meta talk unless the context truly requires it
76+
77+## Examples
78+
79+AI-ish:
80+> Additionally, this feature underscores the importance of robust data validation and fosters a more seamless user experience.
81+
82+Human:
83+> It validates input up front, so bad requests fail early and the UI stops tripping over edge cases.
84+
85+AI-ish:
86+> This library plays a pivotal role in the broader ecosystem.
87+
88+Human:
89+> A lot depends on this library, so small changes here can break other services.
90+
91+AI-ish:
92+> Of course - I'd be happy to help. Here's a breakdown of the issue:
93+
94+Human:
95+> The bug is in the retry loop:
96+
97+AI-ish:
98+> Despite these challenges, the project continues to evolve in response to a dynamic landscape.
99+
100+Human:
101+> The project still works, but these two problems keep slowing it down.
102+
103+## Clarity wins
104+
105+Drop the human-writing style when precision matters more than vibe:
106+- security warnings
107+- destructive commands
108+- legal or compliance text
109+- medical or safety guidance
110+- multi-step procedures where conversational tone could blur the order
111+
112+Write those parts plainly and exactly. Resume normal human-writing style after the critical bit.
113+
114+## Boundaries
115+
116+Do not:
117+- fake personal experience
118+- imitate a demographic, dialect, or identity the user did not ask for
119+- add mistakes to seem human
120+- hide uncertainty behind confident tone
121+
122+Code, commits, and PR text: write normally unless the user asks otherwise.
human-writing/references/research-notes.md+147−0
@@ -0,0 +1,147 @@
1+# Research Notes for `human-writing`
2+
3+This file captures the research basis behind the skill. Keep the
4+main `SKILL.md` practical; keep detailed evidence here.
5+
6+## Wikipedia article and source
7+
8+### Wikipedia: Signs of AI writing
9+Source page:
10+- https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing
11+- Source wikitext: https://en.wikipedia.org/w/index.php?title=Wikipedia:Signs_of_AI_writing&action=raw
12+
13+Key takeaways used in the skill:
14+- The page treats AI tells as **patterns**, not proof. A single word or
15+ one awkward sentence is not enough.
16+- Many strong tells are about **generic significance claims**,
17+ **superficial analysis**, **promotional wording**, **vague
18+ attribution**, and **rigid essay structure**.
19+- It flags repeated "AI vocabulary" clusters such as `additionally`,
20+ `align with`, `delve`, `underscore`, `pivotal`, `vibrant`,
21+ `showcase`, and `tapestry`.
22+- It highlights broader structural habits: rule-of-three phrasing,
23+ overused em dashes, Markdown-heavy output, canned collaborative
24+ phrasing, and generic "challenges / future outlook" conclusions.
25+- It also includes a useful warning section on **ineffective
26+ indicators**: blandness, transition words in isolation, and formal
27+ prose are not reliable proof by themselves.
28+
29+How it shaped the skill:
30+- Focus on removing formulaic phrasing and generic uplift, not on
31+ banning individual words absolutely.
32+- Emphasize specificity, judgment, and natural rhythm over detector
33+ gaming.
34+- Add boundaries so the skill does not encourage false claims of human
35+ authorship.
36+
37+## Research on distinctive AI style
38+
39+### Russell, Karpinska, Iyyer (ACL 2025)
40+Source: https://aclanthology.org/2025.acl-long.267/
41+
42+Why it matters:
43+- Frequent ChatGPT users were strong detectors of AI-generated text and
44+ outperformed many automated detectors.
45+- Their explanations relied not just on word lists, but also on
46+ **formality, originality, and clarity**.
47+
48+How it shaped the skill:
49+- The skill focuses on overall feel and judgment, not only lexical
50+ substitutions.
51+- "Human" writing should sound selective and situated, not just less
52+ detectable.
53+
54+### Juzek and Ward (COLING 2025)
55+Source: https://aclanthology.org/2025.coling-main.426/
56+
57+Why it matters:
58+- Documents lexical overrepresentation in LLM writing, with words such
59+ as `delve`, `intricate`, and `underscore` appearing far more often
60+ after LLM adoption.
61+- Treats this as a measurable linguistic drift, not just a meme.
62+
63+How it shaped the skill:
64+- Include a short watchlist of repeated AI-favorite wording.
65+- Frame the issue as **clusters and repetition**, not one forbidden
66+ token.
67+
68+### Kobak et al. (Science Advances 2025)
69+Source: https://www.science.org/doi/10.1126/sciadv.adt3813
70+
71+Why it matters:
72+- Shows abrupt post-2022 increases in excess vocabulary across more
73+ than 15 million biomedical abstracts.
74+- Reinforces that LLM writing leaves a measurable lexical footprint at
75+ scale.
76+
77+How it shaped the skill:
78+- Prefer plain verbs and concrete nouns over polished but generic
79+ vocabulary.
80+- Warn against abstract noun piles and repetitive stylistic markers.
81+
82+### Rudnicka (Scientific American, 2025)
83+Source: https://www.scientificamerican.com/article/chatgpt-and-gemini-ai-have-uniquely-different-writing-styles
84+
85+Why it matters:
86+- Argues that major chatbots have distinct **idiolects**, much like
87+ humans do.
88+- Implies that AI tells shift over time and differ by model.
89+
90+How it shaped the skill:
91+- Avoid hard-coded, forever bans.
92+- Teach broader habits: specificity, rhythm, audience matching, and
93+ selective judgment.
94+
95+### Saha and Feizi (2025)
96+Source: https://arxiv.org/abs/2502.15666
97+
98+Why it matters:
99+- Detectors often flag even lightly AI-polished human text as AI.
100+- This creates false accusations and shows the limits of detector-first
101+ thinking.
102+
103+How it shaped the skill:
104+- Optimize for better writing, not for beating detectors.
105+- Keep the skill centered on readable, honest prose.
106+
107+## Official style guidance on sounding human without sounding sloppy
108+
109+### Google developer documentation style guide - Voice and tone
110+Source: https://developers.google.com/style/tone
111+
112+Key takeaways used in the skill:
113+- Aim for a voice that is conversational, friendly, respectful, and
114+ natural.
115+- Sound like a **knowledgeable friend**, not a pedant and not a clown.
116+- Avoid buzzwords, placeholder phrases, choppy repetition, and
117+ long-winded sentences.
118+
119+### Digital.gov - Plain Language Web Writing Tips
120+Source: https://digital.gov/resources/plain-language-web-writing-tips
121+
122+Key takeaways used in the skill:
123+- Get to the point quickly.
124+- Use contractions, active voice, familiar words, and short paragraphs.
125+- Put the most important information first.
126+
127+### 18F Content Guide - Be concise
128+Source: https://guides.18f.gov/content-guide/our-approach/be-concise/
129+
130+Key takeaways used in the skill:
131+- Be specific, informative, brisk, and human.
132+- Use contractions and short sentences.
133+- Vary sentence length so writing stays readable without sounding
134+ robotic or clipped.
135+- Avoid adjective-heavy spin and caveat-bloated grammar.
136+
137+## Practical synthesis
138+
139+Across the sources, the same advice keeps showing up:
140+
141+1. Lead with the point.
142+2. Prefer concrete nouns and verbs over abstract uplift.
143+3. Use contractions when natural.
144+4. Vary sentence rhythm.
145+5. Match the audience's register.
146+6. Avoid repeated formulae, hype, and canned summaries.
147+7. Keep clarity and honesty above "sounding human".