Vibe-coded products are ugly in the same specific way: indigo gradients, glowing cards, emoji everywhere, a mascot in every corner, ALL-CAPS stat-cards. It's what a machine reaches for when it has no taste but wants to look impressive, and it's so common you've stopped seeing it.
Kill AI Slop is two things:
- A website, a multilingual field guide (English, Chinese, Japanese, and Korean) that catalogues 33 AI-slop tells, each with an interactive before→after demo showing the clean fix. The site is itself built as a rebuttal to everything it catalogues.
- An Agent Skill,
kill-ai-slop, which turns that catalogue into action: it scans a web project for the code-level signals of each tell, explains why each reads as machine-made, and proposes (or applies) the clean fix.
Every tell is one you can find in real, shipped products, not a strawman. Each entry rebuilds the tell in plain HTML so you can see the machine-default and the clean fix side by side.
website/ Astro site, the field guide (multilingual, static, zero-JS-by-default)
skill/ the kill-ai-slop Agent Skill (SKILL.md + references + scanner)
Live at killaislop.com. To run it locally:
cd website
npm install
npm run dev # http://localhost:4321
npm run build # → dist/ (static, deploy anywhere)Design system: paper + ink, one editorial red used like a proofreader's pen,
hierarchy from scale and space, hairline rules, no gradients / emoji / glass /
badges. See website/src/styles/tokens.css for the rules it holds itself to,
and website/src/data/catalogue.ts for the taxonomy (the single source of truth
shared with the skill).
Run the following command and follow the prompt to install it to your favorite coding agents:
npx skills add yetone/kill-ai-slop
Install it without touching a single file: paste this to your coding agent (Claude Code, Cursor, etc.) and it fetches the skill and drops it into its own skills directory for you.
Install the kill-ai-slop skill from
https://github.com/yetone/kill-ai-slop/tree/main/skill
Copy everything in that directory into a kill-ai-slop/ folder
inside your agent's skills directory, then confirm it's registered.
Then just tell it to "kill the AI slop in this project". Prefer to install by hand?
See skill/README.md.
You can also run the scanner directly. Point it at any web project:
node skill/scripts/scan.mjs path/to/project # grouped report
node skill/scripts/scan.mjs path/to/project --json # machine-readable--only= / --skip= filter by tell id, --exclude= drops paths,
--rules=extra.mjs loads your own language- or stack-specific tells, and
deslop-ignore comment directives suppress hits you've confirmed as
intentional — see skill/README.md.
The scanner is dependency-free and never edits files. In an Agent-Skills host
(e.g. Claude Code), invoke the skill and it will scan, triage, report, then fix,
following skill/SKILL.md. See skill/references/ for the taxonomy, detection
patterns, and remediation playbook, and skill/README.md for install steps.
The demos are reconstructed in plain HTML, for commentary, criticism, and education. The point is not to dunk on any one project. It's to make the ugly defaults visible again, so you can choose against them.