anti-slop-code
Skill9 starsChanged 2 months agoMITView on GitHub
Detect and fix AI-generated frontend implementation slop: isolated helpers that ignore other modules, needless useEffect, type lies, empty wrappers, and boilerplate error handling that misses field edge cases. Trigger for “kode terasa AI”, “rapikan kode”, “dampak ke modul lain”, “edge case”, “error handling generik”, “too many useEffect”, or “over-engineered”. Visual slop stays on anti-slop-design; copy stays on content-design.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- NovusEdge/curttheir own version, first seen Oct 9, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- NovusEdge/curt★ 1same 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 repository36 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .cursor/rules/design-system-enforcer.mdcCursor rule
- frontend-expertSkill
- accessibilitySkill
- anti-slop-designSkill
- app-shellSkill
- architectureSkill
- componentsSkill
- content-designSkill
- data-fetchingSkill
- design-colorSkill
- design-directionSkill
- design-foundationsSkill
- design-surfacesSkill
- design-typographySkill
- design-vocabularySkill
- devtoolsSkill
- engineering-vocabularySkill
- fidelitySkill
- formsSkill
- frontend-expertSkill
- marketingSkill
- monitoringSkill
- motion-inspectionSkill
- motionSkill
- performanceSkill
- polishSkill
- prototypeSkill
- quality-loopSkill
- responsiveSkill
- seoSkill
- skill-authoringSkill
- testingSkill
- tokensSkill
- webglSkill
The file
Files
Anti-slop Code
Overview
Generated code often looks tidy inside one function while ignoring callers, shared cache, server round-trips, and failures seen in the field. Scan blast radius first. Do not treat a visual restyle or a folder rewrite as a code-slop fix.
When to use
- Implementing or reviewing React/FE modules that look generated
- “Rapikan kode” / over-engineered hooks / eslint-disable spray
- A change that may hit other modules, cache, or server load
- Generic
try/catchor toast that skips real failure modes - Cleaning a screen without changing layout, brand, or copy
When to skip
- Purple/gradients/Lorem-as-layout →
anti-slop-design - Generic marketing/UI sentences →
content-design - Public props/
show*soup →components - Feature folders / state ladder / global store →
architecture - Query/loader chrome (skeleton, retry control) →
data-fetching - Measured LCP/INP/CLS campaigns →
performance - Security hardening (XSS, CSP, secrets) — out of this skill
MUST
| Rule | Detail |
| Blast radius | Name callers, sibling modules, shared types, cache keys, and server calls this change touches |
| Not isolated | Do not “fix” one helper by duplicating fetch, invalidating the world, or shifting load onto the client |
| Preserve intended UX | Same product behavior unless the user asked to change it; covering missed failures is required |
| Effects are sync | Derived values render; effects talk to external systems |
| Honest types | No any or as unknown as without a recorded reason |
| Delete first | One-shot wrappers, unused exports, fake setTimeout loading |
| Second-use extract | Don’t add a util/hook/context for a single call site |
| Field errors | Map this surface’s failures; never one generic catch/toast. Unknown modes are labeled, not invented |
Workflow
- Context — callers, shared cache, existing loaders, server round-trips, scale (list size, chatty calls).
- Scan — effects, type lies, wrappers, disable comments, isolated helpers.
- Failures — abort, auth, empty vs 404, conflict, 429, validation, partial, offline, double-submit.
- Handoff — visual/copy/API/folder/fetch-chrome/CWV to their owners.
- Simplify — delete; share existing data; keep project conventions.
- Verify — neighboring modules still compile/behave; loading/error/empty remain; UI unchanged unless asked.
Boundaries
- May decide: inline vs a helper once a second call site exists; keep a disable with a one-line reason.
- Must not: restyle the UI, invent a store, extract a kit, fabricate retry/idempotency, or run a security pass under this name.
Checklist
- Blast radius named (modules / cache / server)
- No new isolated fetch or unbounded client work
- Derived state is not an effect
- No unexplained
any/ type assertion / eslint-disable - Failure modes listed; generic catch replaced or labeled unknown
- Visuals, copy, and folders unchanged unless those skills were in scope
Depth
Full catalog: references/anti-slop-code.md.
Visual catalog: references/anti-patterns.md.
State/folders: references/fe-architecture.md.
Async chrome: references/data-fetching.md.
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
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.