AntiSlop

Scores a draft for AI-writing tells and rewrites the flagged lines in place.

di aplaceforallmystuff · aplaceforallmystuff/the-antislop

Promosso ★ 8.4/10

AntiSlop — Scores a draft for AI-writing tells and rewrites the flagged lines in place.

Cosa fa

Detects AI-writing tells (delve/leverage-style buzzwords, staccato fragment spam, corporate hedging, chatbot sign-offs) across 3 severity tiers, computes a quantified slop score, and rewrites the flagged lines in place with a before/after table. Triggers on requests to check, humanize, or audit text before publishing.

Rapporto di test

Ran it on a paragraph seeded with 12+ tier-1 phrases (delve, leverage, game-changer, moreover...) plus a staccato-fragment run and a chatbot sign-off: it caught and cited essentially every one in a before/after table with a quantified 49-to-6 score, versus a plain edit that cleaned the prose but left no audit trail or count of what was actually wrong.

Testato il: 2026-07-15 · Claude Code 2.x (agent harness)

Installazione

git clone https://github.com/aplaceforallmystuff/the-antislop.git
mkdir -p ~/.claude/skills
cp -r the-antislop ~/.claude/skills/antislop

Comandi e prompt di esempio

  • /antislopScores a draft for AI-writing tells and rewrites the flagged lines in place.

Gli skill si attivano con richieste in linguaggio naturale, senza comandi da ricordare. Dopo l'installazione, prompt come questi lo attivano (in inglese):

  • Does this blog post sound like AI wrote it? Check before I publish
  • Humanize this paragraph, it feels stiff and robotic right now
  • Run a slop check on this draft and fix the flagged lines in place