Skip to content
Back to skills

Slop Sweep

ASecurity

Find unjustified AI-like filler, visual or wording residue, needless abstraction, and process bloat in a candidate; optionally repair bounded presentation-only defects. Use for reviewing interfaces, prose, code, analytics, plans, or explicit slop review.

  • 5 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 19, 2026
ai-agentsrustperformance

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 19, 2026

npx -y skills add simonasrazm/skills --skill slop-sweep --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Slop Sweep?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Slop Sweep
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/simonasrazm-slop-sweep/badge)](https://www.skillsdirectory.com/skills/simonasrazm-slop-sweep)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: slop-sweep
description: Find unjustified AI-like filler, visual or wording residue, needless abstraction, and process bloat in a candidate; optionally repair bounded presentation-only defects. Use for reviewing interfaces, prose, code, analytics, plans, or explicit slop review.
---

# Slop Sweep

Judge residue against the artifact's intent, audience, project conventions, and real cost, not against a blacklist of fashionable words, colors, gradients, cards, or styles. An unconventional trait is not a defect by resemblance alone.

Inspect what was actually produced. Look for generic filler, redundant explanation, obvious generated copy, decorative UI without purpose, duplicated structure, needless abstraction, speculative extensibility, tracker ceremony, and output that cannot justify its maintenance or attention cost. State the demonstrated consequence to clarity, trust, usability, performance, maintenance, or scope.

Repeated long em dashes in generated prose are a common AI-slop signal. Replace them with natural punctuation unless the source or project voice deliberately uses them; judge isolated use in context rather than treating punctuation alone as proof of generation.

Under explicit or delegated fix authority, this skill may repair an unambiguous presentation-only defect whose meaning and behavior remain unchanged: wording cleanup, redundant labels, mechanical token alignment, or an equally local correction. Give the mutation a new candidate identity and obtain a recheck in a fresh context supplied with the accepted contract and candidate, without inherited builder conversation or rationale; this skill does not accept its own repair. The same independence boundary applies when this skill supplies acceptance for another builder.

Route changes to behavior, product meaning, information architecture, design-system rules, code architecture, data, or broad structure to the implementation owner. When invoked as a check only, leave all maintained output unchanged and return evidence plus the smallest repair boundary.

Return findings or the repaired candidate in-channel; create no report or run-state artifact unless requested.

Files in this skill

  • SKILL.md2.1 KB
  • agents/openai.yaml40 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…