the-hooman-loop/anti-slop-kit

the-hooman-loop/anti-slop-kit

Skill Claude CodeCodex

Core judgment guard against generic, unverified AI slop across code, writing, and design. Keeps humans responsible for taste, truth, context, and consequences. Use when generating or reviewing any meaningful artifact, or when the user mentions slop, generic AI output, or "make this less AI.".

the-hooman-loop/anti-slop-kit

Skill Claude CodeCodex

Removes AI-generated code slop (redundant comments, over-engineering, verbose names, boilerplate, dead defensive code) and normalizes style. Use when the user asks to clean up code, remove AI slop, de-slop, fix code style, refactor for readability, or make code more idiomatic.

the-hooman-loop/anti-slop-kit

Skill Claude CodeCodex

Strips the generic AI look from UI and visual design. Subtract-first — one accent, hierarchy from scale and space, reject glow/gradient/emoji pile-on and the default AI aesthetics. Use when the user wants to deslop a UI, make a design less AI, or give an interface a distinctive, intentional look.

the-hooman-loop/anti-slop-kit

Skill Claude CodeCodex

Removes AI writing patterns from prose so it sounds like a specific human wrote it. Two modes — edit (rewrite) or detect (flag patterns without rewriting). Use when the user wants to deslop writing, make text sound human, cut AI tells, or audit a draft for slop.

the-hooman-loop/anti-slop-kit

Skill Claude CodeCodex

Make small, reviewable changes instead of large rewrites. Use when editing an existing codebase or document, so changes stay scoped, preserve existing design and naming, and are easy to review and test.

the-hooman-loop/anti-slop-kit

Skill Claude CodeCodex

Require verification before claiming work is complete. Use before saying something works, is done, or is fixed, and whenever reporting results of code, builds, UI, links, data, or calculations.