Skip to main content
AIPromptIndex
Claude Writing beginner

AI Slop Detector

Added Jul 23, 2026

You are an editor auditing a piece of writing for AI-sounding patterns. Do NOT rewrite anything. Your only job is to find and name each pattern, quote the exact line where it appears, and suggest a fix in a few words. Important: do not guess whether AI wrote this text, and do not give it a score. AI detectors guess; named patterns are evidence the writer can check for themselves. The text to audit: [TEXT] Check for every one of these patterns: 1. Banned words: delve, foster, leverage, utilize, facilitate, empower, streamline, robust, cutting-edge, game changer, tapestry, realm, beacon, multifaceted, meticulous, paramount, transformative, elevate, embark, harness, ever-evolving 2. Empty filler phrases: "it's worth noting," "at the end of the day," "in today's world," "in the age of," "the reality is," "let's dive in" 3. Binary contrasts: "It's not X. It's Y." / "The question isn't X, it's Y." 4. Throat-clearing openers: "Here's the thing," "Let me be clear," "I'll be honest" 5. Faux-insight setups: "What nobody tells you," "The part everyone misses" 6. Dramatic colon reveals: "The best part: it learns." 7. Trailing -ing analysis: "...highlighting," "...underscoring," "...showcasing" 8. Importance puffery: "marks a pivotal moment," "stands as a testament," "plays a vital role" 9. Weasel attribution: "experts agree," "studies show," "widely regarded as" 10. Fake-strong verbs: "serves as," "acts as," "functions as" where "is" would be clearer 11. Synonym cycling: rotating through "the agent... the assistant... the tool" for the same thing 12. Negative listing: "Not a X. Not a Y. A Z." 13. Dramatic fragments: "That's it. That's the whole thing." 14. Robotic rhythm: repeated sentence shapes or stacked punchy fragments 15. Rhetorical setups: "What if I told you...", "Plot twist:", self-answered questions 16. Fake-profound kicker: a final line that turns the point into a cute metaphor or mic-drop 17. Summary-recap ending: "In conclusion," "Ultimately," or a final paragraph restating the piece 18. Formatting slop: emoji in headings, bold sprinkled mid-sentence, bullets that should be prose 19. Em dash overuse: more than 1-2 in a short piece, or clusters used as a rhythm crutch Output format — for each pattern found: PATTERN: [name] QUOTE: "[the exact line from the text]" FIX: [suggested fix in ten words or fewer] End with a one-line count of total patterns found, and offer to edit the draft if I want the fixes applied.
0
Share
Try in Claude

About This Prompt

This prompt audits any piece of writing for the specific, checkable patterns that make text read as AI-generated — without rewriting a word and without pretending to be an AI detector. That distinction matters: AI detection tools guess and produce false positives, while this audit names each pattern it finds, quotes the exact offending line, and suggests a fix you can accept or ignore. It checks nineteen pattern categories, from the banned-word list (delve, leverage, robust) through structural tells like binary contrasts, synonym cycling, fake-profound endings, and em dash overuse. Adapted from Peter Yang's open-source no-ai-slop editing skill (MIT licensed). Useful for auditing your own AI-assisted drafts before publishing, reviewing freelancer submissions, or teaching your team what to watch for.

Variables to Customize

[TEXT]

The text you want audited for AI-sounding patterns

Example: <!--email_off-->In today's ever-evolving digital landscape, this tool isn't just another app. It's a game changer. Experts agree that...<!--/email_off-->

Tips for Best Results

  • Use this before publishing anything drafted with AI help — it catches patterns you have gone blind to after three rounds of edits
  • The audit deliberately never says "this was written by AI" — quoted patterns are evidence you can verify, a verdict is just a guess
  • Zero findings does not mean the writing is good, only that it avoids the common tells; judge the substance separately
  • Pair it with the AI Slop Remover prompt when you want the fixes applied instead of just flagged

Example Output

<!--email_off-->PATTERN: Filler phrase + banned word
QUOTE: "In today's ever-evolving digital landscape"
FIX: Delete the phrase, start with the subject.

PATTERN: Negative listing / binary contrast
QUOTE: "this tool isn't just another app. It's a game changer."
FIX: Say what the tool does instead.

PATTERN: Weasel attribution
QUOTE: "Experts agree that..."
FIX: Name the expert or cut the claim.

TOTAL: 4 patterns found in 2 sentences. Want me to apply these fixes to the full draft?<!--/email_off-->
editing ai-detection ai-slop content-quality proofreading

Get the Best AI Prompts Weekly

Curated prompts, tips, and guides delivered to your inbox every week. Free.

No spam, ever. Unsubscribe anytime.