nais/doc/plain-language
Claude Code and agent skill by nais/doc
- community
Checking that the SKILL.md file is still in the repository.
Edit text for plain, natural English. Use when: reviewing documentation for AI writing patterns, removing corporate jargon, cleaning up generated content, polishing drafts, improving readability. Detects and fixes overused AI phrases, filler words, rhetorical patterns, and structural tells.
Install
npx vskill@latest install nais/doc/plain-languageSKILL.md
Plain Language Editor (English)
Transform text into plain, natural English. Preserve the author's meaning and structure while removing AI-generated markers, jargon, and unnecessary complexity.
When to Use
- After generating or drafting documentation
- When reviewing PRs for AI writing patterns
- When text sounds corporate, robotic, or over-formal
- When asked to "clean up" or "polish" content
What You Do
- Read the text and scan for AI markers listed below
- Make targeted edits to flagged phrases
- Preserve technical terms, numbers, facts, and structure exactly
What You Don't Do
- Add new content or restructure sections
- Rewrite from scratch
- Fix logical issues or factual errors
- Change technical terminology
AI Markers to Fix
Overused phrases (replace or remove)
| AI marker | Fix | |
|---|---|---|
| "delve into", "deep dive" | "examine", "explore", or say what you mean | |
| "navigate" (metaphorical) | "handle", "deal with", "address", or drop it | |
| "at the heart of", "at its core" | Cut; start with the actual subject | |
| "landscape", "ecosystem", "paradigm" | Concrete noun: "market", "field", "approach" | |
| "robust", "holistic", "nuanced", "comprehensive" | Cut or use specifics: "reliable", "complete", "detailed" | |
| "seamless", "streamlined", "optimized" | Say what actually improved, or cut | |
| "leverage" (as verb) | "use" | |
| "tapestry", "mosaic", "fabric", "kaleidoscope" | Cut the metaphor; "variety" or "range" if needed | |
| "realm", "arena", "space" (metaphorical) | "field", "area", or name the thing | |
| "toolkit", "roadmap", "arsenal" | "tools", "plan", "options" | |
| "journey" (metaphorical) | "process", "progress", or cut | |
| "multifaceted", "complex interplay", "granular" | "several factors", "interaction", "detailed" | |
| "empower", "enable", "unlock", "unleash", "amplify" | Say what concretely happens | |
| "foster", "cultivate", "drive", "facilitate" | "build", "create", "improve" | |
| "shed light on", "underscore", "showcase" | Direct statement | |
| "at scale", "end-to-end", "best practices" | Keep only if technically precise; otherwise cut | |
| "game changer", "paradigm shift", "thought leadership" | Replace with specific impact or drop | |
| "cutting-edge", "next-generation", "revolutionary", "transformative", "innovative", "groundbreaking" | Say what's actually new | |
| "synergy", "bandwidth", "deliverables", "stakeholders" | "cooperation", "time", "results", "people" | |
| "harness", "embark", "strive", "elevate" | "use", "start", "try", "improve" | |
| "unpack", "tap into" | Say what you mean directly | |
| "at the forefront of", "pushing the boundaries" | Say what's new or different | |
| "cornerstone", "linchpin", "bedrock" | Name the dependency or say why it matters | |
| "a testament to", "serves as a reminder" | Cut the frame; state the thing directly |
Filler precision words (cut or replace)
These sound precise but add nothing. AI uses them 5-10x more than human writers:
| AI marker | Fix | |
|---|---|---|
| "significant", "significantly" | Say how much, or cut | |
| "notable", "notably" | Just state the thing | |
| "meaningful" (as in "meaningful impact") | Say what the impact is | |
| "considerable", "substantial" | Use a number or cut | |
| "crucial", "critical", "vital", "essential", "key" | One per article is fine. Five is a pattern | |
| "impactful" | Say what the effect was |
Needlessly formal wording
AI defaults to stuffy vocabulary. Simplify when the formal word adds no precision over the plain alternative. Keep technical terms when they are standard in the domain (e.g., "implement" is fine in code docs).
| AI marker | Fix | |
|---|---|---|
| "utilize" | "use" | |
| "commence" | "start", "begin" | |
| "endeavor" | "try", "effort" | |
| "paramount" | Say why it matters, or "important" | |
| "aforementioned" | Name it, or "this", "that" | |
| "subsequently" | "then", "after" | |
| "pertaining to" | "about" | |
| "necessitate" | "need", "require" | |
| "elucidate" | "explain" | |
| "plethora", "myriad" | "many", "plenty" | |
| "optimal" (non-technical) | "best" | |
| "ascertain" | "find out", "check" | |
| "disseminate" | "share" | |
| "meticulously" | "carefully", or say how |
Sentence-initial adverb stuffing
Flag density (more than 2-3 per page):
- "Importantly," -- Cut. If it's in the article, the reader knows it matters.
- "Notably," -- Cut. Just state it.
- "Interestingly," -- Cut. Let the reader decide.
- "Significantly," -- Cut or use a number.
- "Crucially," -- Cut.
- "Ultimately," -- Usually a signpost for a conclusion you haven't earned. Cut or move the claim to the start.
2026 patterns
| AI marker | Fix | |
|---|---|---|
| "quiet confidence", "quiet rebellion", "quietly growing" | Rewrite without "quiet" | |
| "in today's fast-paced world", "in an increasingly X world" | Delete entirely | |
| "in the digital age", "as technology advances" | Delete or name what changed | |
| "X has become an integral part of our lives" | Cut the boilerplate | |
| "genuinely", "truly", "actually", "remarkably" | Cut | |
| "arguably", "typically", "generally speaking", "broadly speaking" | Commit to the claim or acknowledge specific uncertainty | |
| "undeniable", "undoubtedly", "invaluable", "pivotal" | Replace with evidence or specifics | |
| "It's not about X, it's about Y" (repeated pattern) | Vary the construction; use once at most | |
| "No X. No Y. Just Z." | Formulaic emphasis. Rewrite or vary | |
| "According to recent studies" (without citation) | Cite it or cut the appeal to authority |
Empty authority frames
These add a veneer of objectivity without substance. Remove the frame and keep the claim only if it stands on its own.
| AI marker | Fix | |
|---|---|---|
| "It is widely acknowledged that..." | By whom? Cite or just state the claim | |
| "There is growing evidence that..." | Cite the evidence or cut | |
| "Research suggests that..." (uncited) | Name the research or cut | |
| "A common misconception is that..." | Just state the correct thing | |
| "It is essential to strike a balance" | Say between what, and why | |
| "While some may argue X, others believe Y" (with no source for either) | Take a position or cite both sides |
Faux intimacy phrases (cut)
| AI marker | Fix | |
|---|---|---|
| "Here's the thing", "Here's the kicker", "But here's the thing" | Just state the thing | |
| "Here's an uncomfortable truth", "Here's what most people miss" | Skip the buildup; say it directly | |
| "Honestly?" (followed by text) | Delete "Honestly?" | |
| "The best part?" | Cut the hype; just explain | |
| "You're not alone", "You're not imagining it" | Delete unless actually offering comfort |
Signposting phrases (remove)
- "It's worth noting that", "It's important to remember" -- Just state the thing
- "This means several things", "This suggests that" -- Trust the reader; cut the signpost
- "To put it simply", "Simply put", "In essence", "Fundamentally" -- Just say it
- "A key takeaway is", "From a broader perspective" -- Cut the meta-commentary
- "Moving forward", "Going forward", "The road ahead" -- Use specific time references or delete
Transition word overload
- "Moreover", "Furthermore", "Consequently", "Thus", "Hence" -- Use sparingly. Replace most with "also" or connect naturally
- "Nevertheless", "Notwithstanding", "Accordingly" -- Replace with "But", "Still", "So"
- "That being said", "On the other hand" -- Vary or remove
- "In addition to", "As a result", "Therefore" -- Fine occasionally; flag if every paragraph
Rhetorical patterns (restructure)
- Tyranny of three: Enumerations always in threes? Vary the count.
- Perfect antithesis: "Not just X, but Y" -- use sparingly.
- Fake questions: "How do we solve this?" followed by the answer -- cut the question.
- Excessive hedging: "often", "typically", "generally", "can be", "may" -- commit where you can.
- Relentless balance: Both sides of every point without consequence -- take a position.
- LLM-safe truths: "consistency is important" -- cut the filler.
- Summary restating: Repeating the point with different words. Cut it.
- "Let's" framing: "Let's explore", "Let's break this down" -- just present the information.
Punctuation and formatting tells
- Em dash overuse: More than 3-4 per 500 words is a pattern. Replace most with commas, parentheses, or periods.
- Semicolons: Replace with periods or commas unless the connection genuinely needs it.
- Colon in every heading and bullet: Vary punctuation.
- Exclamation marks in technical text: Cut them.
Structural tells
- Relentless balance: Sections of identical length. Suggest varying depth.
- Generic specificity: Examples that feel specific without naming real cases. Add concrete details.
- Sentence uniformity: Same beat, same length. Vary structure and rhythm.
- Too tidy: No rough edges. Professional writing has moments of emphasis.
- Compulsive over-explanation: Restating simple points in different words within the same passage. Flag and cut the restatement.
- List addiction: Numbered or bulleted lists for simple, linear ideas that read better as prose. Flag density.
- Formulaic scaffolding: Every section follows intro-context-solution-summary. Flag when the template shows.
- Overly polished grammar: No contractions, no sentence fragments, no informal asides — everything scrubbed to an unnatural shine. Flag in non-reference prose.
- Repetitive paragraph openers: Consecutive paragraphs starting with the same construction (e.g., "This means...", "This allows...", "This ensures...").
Procedure
When given text to review:
- Scan for the markers above
- Report a summary: how many markers, which categories
- Rewrite with markers removed or replaced
- Show key changes: original phrase to replacement
When editing a file directly:
- Read the file or section
- Make targeted edits to flagged phrases, not surrounding prose
- Preserve technical accuracy and data references exactly
Rules
- Never change technical terms, numbers, or specific facts
- Never add content, only simplify and clarify existing text
- Preserve the author's argument structure: fix the surface, not the skeleton
- Context matters: flag density, not individual occurrences
- Don't over-colloquialize: match the intended register
- Vary your edits: don't create new patterns by using the same replacement every time
- Markers are signals, not banned words: judge by density, vagueness, and surrounding context
- Structural tells are detection-only: flag them in reports but don't restructure sections
Detection Tips
- The 8-10 rule: 8-10 flagged phrases in a single piece = high AI probability
- Read aloud test: Sounds like a textbook = AI
- Tricolon count: Lists always in threes = AI
- Emotional temperature: Flat throughout = AI; spikes of conviction = human
- Low burstiness: Uniform sentence and paragraph length throughout = AI. Human writing has natural rhythm variation.
- Impossible to disagree with: Banal statements that don't advance the point ("Consistency is important") = filler. Cut or make specific.
- No named examples: Generic scenarios instead of real tools, real errors, real configurations = AI pattern
- Density, not instances: These are markers, not banned words. Judge by density, vagueness, and context. One "crucial" is fine; five in a page is a pattern.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryChecking.github/skills/plain-language/SKILL.md
Popularity
- Repository stars
- 68
- Repository forks
- 62
- 7-day trend score
- 21
Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 25 Aug 2026, 2 months ago.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/nais/doc/plain-language)