Skip to content
SkillModel-led12,805 tokens in use

Antislop

The core filter of the antislop rulebook: 40 rules, a liveliness toolkit and a delivery gate that stop generic AI-made interfaces.

9,688 stars on the repository MIT Licensemiqdadbadjuber/anti-slop · skills/antislop
npx skills add miqdadbadjuber/anti-slop
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing Antislop costs

26tokens idle
12,805tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

shelldetectedwrites filesdeclared

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Good
  • Frontmatter8/15
  • Structure25/25
  • Examples6/15
  • Output15/15
  • Size9/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Good
  • Development activity100
  • Maintenance100
  • Community52
  • Maturity20
  • Momentum40

Antislop is the always-loaded core of the Anti Slop rulebook for AI coding agents. It does not prescribe colors, fonts or layouts. It holds every visual decision to a purpose test, so a technique stays only when you can write down what it serves.

It runs in one of two modes: DURING, where the rules apply while the agent builds, or AFTER, where the agent audits finished work into a numbered findings list you approve. Either way the work ends at a pass or fail Delivery Gate.

What You Get

  • A rulebook the agent applies while it builds UI, so generic patterns do not appear
  • An audit mode that writes a numbered findings list with the rule broken and a priority for each item, and changes nothing until you approve numbers
  • Three liveliness dials (energy, rhythm, motion) set from a Design Read before generation
  • A mandatory pass or fail Delivery Gate in four blocks
  • Hard rules on contrast, keyboard use, UI states, real content and no invented statistics or testimonials

Common Use Cases

  • Building a landing page or app screen that does not look like every other AI page
  • Auditing a finished interface and fixing only the findings you approve
  • Keeping a design system file and an agent in step through a shared rule set
  • Checking a build for accessibility and mobile problems before delivery

What Makes It Unique

Most design prompts add style. This one removes unexplained technique and then requires life to be added back through dials and an identity motif, because a filter alone leaves a sterile page. Its audit mode is approval based: the agent lists numbered findings, each tied to a rule and a priority, and edits only the numbers you pick. The sibling skills for copy, code comments, accessibility, mobile and UI load only when a task needs them.

Under The Hood

The skill is one long SKILL.md plus a VERSION file, with no scripts of its own. It allows the agent to read, write and edit files and to search. It reads a small settings file in your home config folder to learn a saved mode, and treats it as data only. In audit mode it writes a numbered audit file in an anti-slop folder in your project and, after you approve findings, edits the project files. It only writes the settings file when you ask it to remember a mode. It names npx commands for installing or updating the rulebook, but the instructions only describe them and tell the agent to answer questions about updating. It does not fetch web pages or call APIs.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers