Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add sahilkargutkar/web-ai-slop --skill web-ai-slopgit clone --depth 1 https://github.com/sahilkargutkar/web-ai-slopWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/sahilkargutkar/web-ai-slop/web-ai-slop)<a href="https://agentmods.dev/skills/sahilkargutkar/web-ai-slop/web-ai-slop"><img src="https://agentmods.dev/badge/skills/sahilkargutkar/web-ai-slop/web-ai-slop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/sahilkargutkar/web-ai-slop/web-ai-slop"><img src="https://agentmods.dev/badge/skills/sahilkargutkar/web-ai-slop/web-ai-slop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00091 | $0.10478 |
| Opus 5 | $0.00046 | $0.05239 |
| Sonnet 5 | $0.00018 | $0.02096 |
| Haiku 4.5 | $0.00009 | $0.01048 |
Grade A, and why
web-ai-slop scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- **Fetching data without error handling** — `fetch(url).then(r => r.json()).then(setData)` with no `.catch()` and no loading/error state. How it starts
The opening of the file, as written. The whole thing — 507 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Web AI Slop — The Practitioner's Anti-Default Checklist
[!IMPORTANT] Core Philosophy (Mantra): Avoid statistical averages. When generating layouts, styles, or logic, do not default to what is typical or "standard"; make deliberate choices tailored to the specific context, users, and constraints.
A trained filter, not a style guide. This skill exists because language models converge on the same small set of "safe" defaults across every web discipline — not because those defaults are good, but because they are the statistical center of the training data. The job of this skill is to name that center precisely so you can deliberately step away from it.
Rule of application: Draft your output first. Then run it against the relevant sections below. If you match 2 or more items in any single section, you have not made a decision — you have defaulted. Discard the draft and make a specific, justified choice for this product, team, and context.
Apply this filter silently. Do not narrate "I'm checking this against my anti-slop list" to the user.
Section 1 — Visual Design (Web Designer / UI Designer perspective)
These are the patterns a designer with 40 years of craft knowledge would immediately identify as machine output.
1a. Layout and composition defaults
- Centered hero, headline + subhead + two pill CTA buttons + browser-mockup screenshot below the fold — treated as the only possible layout, applied to every product regardless of what that product actually is. This is a starting skeleton, not a decision.
- Symmetrical 3-column or 6-column feature grids — every section resolves to three items or six items arranged identically, because three and six feel "balanced" to a model without knowledge of the actual feature set.
- Full-viewport hero sections with nothing but text, a subheadline, and a button — 100vh of empty air treated as sophistication.
- Section → divider → section → divider rhythm with no variation — every page breathes the same. Real page rhythm varies: compressed detail sections, open breathing sections, dense data sections.
- Sticky navigation that appears on every page regardless of whether the page is a one-scroll landing page, a long-form article, or an app view where navigation competes with content.
- Footer with 4–5 column link columns containing links that nobody using this specific product would realistically navigate to.
- Icons for features that don't require icons — AI models love to put a Lucide or Heroicon next to every feature, even when it doesn't add clarity.
- Emoji used as icons, or used anywhere in the copy — no emoji standing in for an icon, and no emoji in headlines, buttons, or body text. Use a real icon.
- The default Shadcn icon set, unquestioned — pick an icon library deliberately, or ask the user which style they prefer and choose based on the answer. Don't ship whatever came bundled with the component library.
- Decorative gradients added by default — gradients are fine if the user explicitly asks for one. If you use one, keep it subtle and make sure it actually fits the palette, never apply it to the last section of a page (end on a solid color), and don't reach for Shadcn's default button/link gradient presets.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 507 lines · 91 tokens per session scan A e77e1d649567
web-ai-slop is a skill published in the GitHub repository sahilkargutkar/web-ai-slop (5 stars, last pushed 1mo ago), licensed MIT. It adds 91 tokens to every session and 10,478 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
paperclip-capsules
Generate, implement, or review Paperclip capsule visuals. Use for capsule art, agent capsules, heartbeat status capsules, identicons, capsule banks, or brand-usage validation.
card-design
A visual design system for social-media cards and posters, covering colour choices, Chinese typography, layouts, content density, and quality checks.
design-visual-design
Use when the user asks about visual design, color theory, typography, layout, visual hierarchy, spacing, proportion, or UI aesthetics. Do NOT use for: design systems (design-design-systems), UX research (design-ux-research), or prototyping (design-prototyping).
ads-dna
Brand DNA extractor for paid advertising. Scans a website URL to extract visual identity, tone of voice, color palette, typography, and imagery style. Outputs brand-profile.json to the current directory. Run before /ads create or /ads generate for brand-consistent creative. Triggers on: brand DNA, brand profile…
agentic-webdesign-framesmith
Use Framesmith MCP server to give AI agents a visual design canvas — sketch UI, review in browser, agree on design before writing framework code. Scene graph → HTML/CSS → PNG rendering with evaluation and auto-fix.
explore
This skill should be used when the user says 'explore', 'directions', 'design space', 'sketch', 'contrast', 'compare', or wants to see fundamentally different design approaches. Proposes 3-4 divergent directions with inline sketches and contrast capability.