human-writing
Write like a human, not a language model. Use this skill whenever the user asks you to write or rewrite text that should sound human-written — emails, documents, articles, blog posts, reports, bios, descriptions, social media posts, or any prose. Triggers on phrases like "make it sound human", "write this as a human would", "human-writing", "make this more natural", "less AI-sounding", "rewrite this naturally", "sound more authentic", "doesn't sound like AI", "write like a person", or when the user asks you to write anything and adds that it should sound human or natural. Also use when reviewing existing text for AI tells. If the user is asking you to write ANY text and mentions wanting it to sound human, natural, or authentic, use this skill. When in doubt, use it — undertriggering is worse than overtriggering.
What this skill does
# Human writing
You are producing text that should read as if written by a knowledgeable, specific human being — not assembled from high-frequency training tokens. This matters because AI-generated text has a recognizable fingerprint: it regresses toward statistically common language, replaces specific facts with generic positive descriptions, and inserts formulaic structures that recur regardless of topic. The result reads like someone shouting louder and louder that a portrait shows a uniquely important person, while the portrait itself fades from a sharp photograph into a blurry, generic sketch.
Your job is to produce the sharp photograph.
## How to use this skill
**Rewriting existing text:** Read the text, identify AI tells (vocabulary, structure, formatting), then rewrite to eliminate them while preserving the factual content and intent. Present the rewritten version.
**Writing new text:** Apply all the rules below from the start. Write the text once, correctly.
**Reviewing text:** When asked to review, flag specific AI tells with line references and explain what makes each one detectable.
## The core problem with AI writing
Language models regress to the mean. They produce the most statistically likely phrasing that applies to the widest variety of cases. This creates two simultaneous effects:
1. **Specifics disappear.** The precise detail "inventor of the first train-coupling device" becomes the vague "a revolutionary titan of industry."
2. **Emphasis inflates.** Everything becomes important, pivotal, groundbreaking, or a testament to something.
The fix is simple: be specific, be plain, stop editorializing.
---
## Banned vocabulary
These words and phrases are AI tells when used in the ways described. Some are fine in narrow literal contexts (e.g., "underscore" meaning a literal underline character), but their figurative or filler uses are dead giveaways.
**Puffery and promotional words** — never use these in descriptive prose:
nestled, breathtaking, vibrant, boasts (meaning "has"), groundbreaking, renowned, showcasing, captivating, stunning, profound, exemplifies, enhancing, diverse array, rich history, natural beauty, fascinating glimpse, diverse tapestry, dependable value-driven experiences
**Importance inflation** — never attach these to mundane subjects:
testament, pivotal, crucial, vital, significant (as filler adjective), indelible mark, deeply rooted, key turning point, evolving landscape, focal point, enduring legacy/impact/influence
**Filler verbs** — use "is", "are", "has" instead:
serves as, stands as, marks as, represents a, boasts, features (meaning "has"), offers (meaning "has")
**Trailing -ing commentary** — never append these participial phrases:
highlighting its significance, emphasizing the importance, reflecting the rich culture, underscoring its role, fostering a sense of, ensuring, cultivating, demonstrating the ongoing relevance, confirming its relevance, illustrating lasting influence, creating a space where, encompassing, contributing to the
**Sentence-starting fillers:**
Additionally (at start of sentence), Furthermore, Moreover (when used as pure transition filler)
**Abstract buzzwords** — avoid clustering these:
delve, landscape (abstract), tapestry (abstract), interplay, intricate/intricacies, meticulous/meticulously, garner, bolstered, underscore (verb), showcase, align with, resonate with, valuable insights, enhance, foster, enduring, key (as adjective)
**Meta-commentary and chatbot tells:**
I hope this helps, Certainly!, Of course!, Would you like, Let me know, Here is a, More detailed breakdown, You're absolutely right, As an AI language model
**Knowledge-gap hedging:**
As of my last knowledge update, Based on available information, While specific details are limited, Not widely documented/available, In the provided sources
For the full linter regex patterns, read `${CLAUDE_PLUGIN_ROOT}/references/detection-patterns.md`.
---
## Banned structures
### The significance formula
Never write sentences that inflate the importance of ordinary facts.
Bad: "The institute was established in 1989, marking a pivotal moment in the evolution of regional statistics."
Good: "The institute was established in 1989."
### The challenges-and-future formula
Never write "Despite its [positive word], [subject] faces challenges" followed by vague optimism. Never create standalone "Challenges and Future Directions" or "Future Outlook" sections. If challenges exist, integrate specific ones into the relevant paragraph.
Bad: "Despite its success in renewable energy, the company faces several challenges in the evolving market landscape. However, ongoing initiatives position it well for future growth."
Good: "The company's wind farms produced 400 MW in 2024, though grid integration costs rose 15% following mandatory infrastructure upgrades under the 2023 federal transmission rules."
### Negative parallelisms
Never use "not only...but also", "not just...it's", "it's not about...it's", or "no...no...just" constructions to appear balanced.
Bad: "It's not just a meme — it's a celebration of grassroots car culture."
Good: State what it is, directly.
### Ritual conclusions
Never end sections with "In summary", "In conclusion", "Overall", or restatements of what the section already said. Stop when the last fact is stated.
### Vague attributions
Never write "industry reports", "experts argue", "some critics argue", "researchers and conservationists", or "several sources/publications" when you mean one or two specific sources. Name the source or don't make the claim.
Bad: "Researchers have noted promising developments."
Good: "Chen et al. (2024) found a 12% efficiency gain."
### The notability pile-on
Never list media outlets that covered a subject just to prove the subject matters. Don't write "featured in Vogue, Wired, Toronto Star, and other media" as proof of importance.
### Ecosystem/heritage inflation
When writing about places, species, organizations, or cultural topics, do not insert generic statements about broader significance, ecosystems, cultural heritage, or conservation unless the user specifically requested that framing.
Bad: "It plays a role in the ecosystem and contributes to Hawaii's rich cultural heritage."
Good: State what the thing actually does or is, with specifics.
---
## Banned formatting
- **No title-case headings.** Use sentence case. "Early career" not "Early Career and Professional Development."
- **No emoji in headings or bullets.**
- **No excessive boldface.** Bold only terms being defined. Not every proper noun or key phrase.
- **No inline-header vertical lists.** Don't format bullets as "**Bold header:** description text." Use plain bullets or prose.
- **No em-dash overuse.** Em dashes are fine occasionally, but AI writing uses them constantly for rhetorical punch. Use commas, parentheses, or colons instead when those are more natural. One em dash per paragraph is a rough ceiling.
- **No unnecessary tables.** If the information fits in one or two sentences, write sentences.
- **No thematic breaks (---) before every heading.**
---
## What to do instead
1. **State facts without editorializing their significance.** Let the reader decide what's important. "The gallery has four rooms" not "The gallery features four distinct spaces that serve as a hub for contemporary expression."
2. **Use specific, concrete details.** Numbers, dates, names, measurements. "The station has six platforms and eight tracks" not "serving as a major hub facilitating the movement of passengers and goods."
3. **Attribute claims to named sources.** "Chen et al. (2024) found a 12% efficiency gain" not "researchers have noted promising developments."
4. **End sections with content, not summaries.** The last sentence should be a fact, not a restatement.
5. **Use simple verbs.** "The gallery has four rooms" not "The gallery boasts four distinct spaces." Prefer "is", "are", "has", "was" over elaborate substitRelated in Image & Video
watch
IncludedWatch a video (URL or local path). Downloads with yt-dlp, extracts auto-scaled frames with ffmpeg, pulls the transcript from captions (or Whisper API fallback), and hands the result to Claude so it can answer questions about what's in the video.
physical-ai-defect-image-generation
IncludedUse when the user wants to orchestrate defect image generation, run associated setup, or handle outputs on OSMO. The Day 0 path handles cold-start with USD-to-ROI, image-edit augmentation, and AnomalyGen to create initial PCBA datasets. The Day 1 path performs inference and labeling on real images. This skill helps with first-time asset setup, creation of finetuning checkpoints, and configuring deployment. Trigger keywords: defect image generation, dig workflow, dig pipeline, defect image detection workflow, aoi pipeline, aoi anomalygen, usd2roi anomalygen, day 0 pcba, day 1 pcba, day 1 real-photo alignment, day 1 manual roi, metal surface anomaly, glass defect, anomalygen finetune, setup_pcb, setup_metal, setup_glass, setup_pretrained, dig setup, dig datasets, dig pretrained checkpoint, dig image-edit endpoint.
accelint-react-best-practices
IncludedReact performance optimization and best practices. ALWAYS use this skill when working with any React code - writing components, hooks, JSX; refactoring; optimizing re-renders, memoization, state management; reviewing for performance; fixing hydration mismatches; debugging infinite re-renders, stale closures, input focus loss, animations restarting; preventing remounting; implementing transitions, lazy initialization, effect dependencies. Even simple React tasks benefit from these patterns. Covers React 19+ (useEffectEvent, Activity, ref props). Triggers - useEffect, useState, useMemo, useCallback, memo, inline components, nested components, components inside components, re-render, performance, hydration, SSR, Next.js, useDeferredValue, combined hooks.
elevenlabs-agents
IncludedBuild conversational AI voice agents with ElevenLabs Platform using React, JavaScript, React Native, or Swift SDKs. Configure agents, tools (client/server/MCP), RAG knowledge bases, multi-voice, and Scribe real-time STT. Use when: building voice chat interfaces, implementing AI phone agents with Twilio, configuring agent workflows or tools, adding RAG knowledge bases, testing with CLI "agents as code", or troubleshooting deprecated @11labs packages, Android audio cutoff, CSP violations, dynamic variables, or WebRTC config. Keywords: ElevenLabs Agents, ElevenLabs voice agents, AI voice agents, conversational AI, @elevenlabs/react, @elevenlabs/client, @elevenlabs/react-native, @elevenlabs/elevenlabs-js, @elevenlabs/agents-cli, elevenlabs SDK, voice AI, TTS, text-to-speech, ASR, speech recognition, turn-taking model, WebRTC voice, WebSocket voice, ElevenLabs conversation, agent system prompt, agent tools, agent knowledge base, RAG voice agents, multi-voice agents, pronunciation dictionary, voice speed control, elevenlabs scribe, @11labs deprecated, Android audio cutoff, CSP violation elevenlabs, dynamic variables elevenlabs, case-sensitive tool names, webhook authentication
humanizer
IncludedHumanize AI-generated text by detecting and removing patterns typical of LLM output. Rewrites text to sound natural, specific, and human. Uses 28 pattern detectors, 560+ AI vocabulary terms across 3 tiers, and statistical analysis (burstiness, type-token ratio, readability) for comprehensive detection. Use when asked to humanize text, de-AI writing, make content sound more natural/human, review writing for AI patterns, score text for AI detection, or improve AI-generated drafts. Covers content, language, style, communication, and filler categories.
generating-mermaid-diagrams
IncludedSalesforce architecture diagrams using Mermaid with ASCII fallback. Use this skill when generating text-based diagrams for Salesforce architecture, OAuth flows, ERDs, integration sequences, or Agentforce structure. TRIGGER when: user says "diagram", "visualize", "ERD", or asks for sequence diagrams, flowcharts, class diagrams, or architecture visualizations in Mermaid. DO NOT TRIGGER when: user wants PNG/SVG image output (use generating-visual-diagrams), or asks about non-Salesforce systems.