agentleFS
samigimansamigiman/anti-slop-skills/skills/deslop-ui/SKILL.md

deslop-ui

Skill0 starsChanged 19 days agoMITView on GitHub

Finds and removes "AI slop" design in an existing site, app, admin panel or HTML/CSS/React/Tailwind codebase — emoji icons, purple/indigo gradients, gradient text, glassmorphism, rounded-3xl and shadow-2xl everywhere, invented stats/testimonials, template heroes, fake countdowns, scroll-reveal, pointless animation, cluttered layouts. Scanner script + checklist + fix recipes. Triggers: AI slop, deslop, clean up the design, design audit, UI audit, UX audit, looks AI-generated, looks generic, looks like a template, cluttered, outdated, tidy up.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. devdotfast/whiteboard★ 3.5ksame name
    2. kmaida/deslop-skills★ 14same name
    3. KyaniteLabs/tastecheck★ 12same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository2 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If a headless or hidden browser pauses animations, sections behind scroll-reveal will look blank — that itself is a finding.

    line 30
  2. Very specific:

    Don't touch copy (use deslop-copy separately, with permission).

    line 43

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Deslop UI — clean up an existing design

For new designs use anti-slop-design. This skill cleans existing code and screens. Functionality, copy and data stay unchanged — only the slop goes, unless the user says otherwise.

Workflow

1. Scope

Which folder, page or component. Read AGENTS.md / CLAUDE.md / DESIGN*.md if present — approved tokens and bans there override this skill. Respect file ownership/locks.

2. Automated scan

python3 -I <skills-dir>/deslop-ui/scripts/slop_scan.py <dir> [--json] [--min P2] [--fail-on P1]

Skips node_modules, dist, .next, build, vendor and minified files. Each hit: file:line, category, severity, fix hint. Hits are signals, not verdicts — check context (a colour swatch named "purple" is fine; backdrop-blur on a sticky header may be fine). Expect the scanner to miss most layout-level slop; step 3 is where the real findings come from.

3. Visual review (mandatory)

Open the pages in a browser at 1440 and 390px and look at the screenshots yourself. Use the "Visual" questions in references/slop-catalog.md: everything centred, the same three-card rhythm, empty hero space, a weak primary action, repeated photos, mixed alignment, nothing product visible on the first mobile screen.

Measure what screenshots can't show reliably — via the browser console:

  • distinct border radii, font sizes, font families and shadow recipes in use (a design system has few of each);
  • elements with a gradient background;
  • clickable elements without cursor:pointer; touch targets under 40px;
  • horizontal overflow; sticky header height on mobile.

If a headless or hidden browser pauses animations, sections behind scroll-reveal will look blank — that itself is a finding.

4. Report (before changing anything)

Table: # | P | Where (file:line or screen area) | What you see | Fix.

  • P1 misleads or confuses the user, or screams "AI-made" (fake urgency, invented numbers, hidden content, no product on first mobile screen).
  • P2 cheap/template look.
  • P3 polish.

Also list what is good ("don't change") so working parts aren't broken, and what you could not verify. If the user only asked for an audit, stop here.

5. Clean up

Follow references/fix-recipes.md:

  • Fix the token layer first (colour, radius, shadow, type) — components follow.
  • Don't touch copy (use deslop-copy separately, with permission).
  • Invented numbers/testimonials/logos: list them for the user before deleting — they might be real.
  • One step = one category, so diffs stay reviewable.

6. Re-verify

Re-run the scanner (P1 = 0), take 1440/390 screenshots, check the console is clean and nothing overflows. Show before/after. State what was not verified.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.