de-slopify
Made by JeffRemove telltale signs of AI-generated "slop" writing. Use when polishing READMEs, docs, marketing copy, or public-facing text, or when text "sounds like ChatGPT".
Works With
Quality and Utility come from canonical skill metadata. Community ratings appear separately as stars when subscriber reviews exist.
Subscribe to download
Public previews are available, but downloads, CLI sync, and premium content require an active subscription.
Sign In to Subscribe- Full skill content with deterministic versioning
- CLI sync + local search index
- Access to Jeffrey’s premium library
Save to your account
Saved skills live in your cloud list. Run jsm sync on each machine to pull them locally.
What's New with this Skill
2Updates, enhancements, and release notes for this skill.
Updated: de-slopify, from six tells to a full field catalog
The de-slopify skill grew from six patterns to roughly two dozen, mined from editor-maintained lists of AI writing signs. It now names the mechanism behind slop (text gets less specific and more exaggerated at the same time), splits fix-on-sight tells from judgment calls, lists red herrings that should never be "fixed", and scopes emdash removal to prose punctuation so dashes in math formulas and code are left alone.
- Setup-reversal copy is the top tell: "Four steps, and only one of them is yours." The fix rule: if the reversal is carrying the sentence, the sentence has no content; find the content and state it
- Copula-avoidance table maps "serves as", "stands as", "marks a", "boasts", and "holds the distinction of being" back to plain "is" and "has"
- AI-vocabulary list of about thirty words judged by density and co-occurrence, never single hits; new entries include spine, seam, and smoke-test
- Trailing "-ing" significance clauses ("...contributing to the socio-economic development of the region") get cut; the sentence almost always survives intact
- Filler gets two operational tests: does the paragraph's content fit in one sentence, and can you state the author's claim halfway through
- Red-herrings section: perfect grammar, absent contractions, a single academic word, and AI-detector output are not evidence; detectors false-positive constantly
- Emdash scope corrected: prose punctuation only; dashes in math formulas, code, CLI flags, and numeric ranges are off limits
- Bold-label bullets (a bolded "Blazing Fast:" label in front of the text) called out as the most recognizable AI structure in READMEs; convert to prose or plain bullets
The original skill covered emdash overuse, "It's not X, it's Y", "Here's why", forced enthusiasm, and empty hedges. Those all survive, and the update adds the patterns that give current model output away: setup-reversal marketing copy ("Ten features. Zero headaches."), copula avoidance ("serves as" where "is" belongs), AI-vocabulary clusters (delve, pivotal, tapestry, spine, seam, smoke-test), trailing "-ing" significance clauses, bold-label bullet lists, the "Despite challenges... future outlook" closing formula, generic analogies, and filler.
Enhanced: De-Slopify, renamed from docs-de-slopify
Previously docs-de-slopify, this skill has been renamed. It removes telltale signs of AI-generated "slop" from documentation and other public text: emdash overuse, "It's not X, it's Y" formulas, "Here's why" lead-ins, and similar LLM tells.
- Targets specific AI writing tells: emdash overuse, "It's not X, it's Y", "Here's why", "Let's dive in", "At its core...", "It's worth noting..."
- Manual, line-by-line review: the skill insists this can't be done with regex or a script
- Emdash alternatives table: semicolons, commas, or splitting into two sentences
- Renamed from docs-de-slopify because the docs- prefix was redundant; works on READMEs, API docs, and any public-facing text