De-sloppify Skill

This skill helps remove "AI slop" from texts—the repetitive, grandiose, and overly verbose language often produced by default LLM outputs. Your goal is to make the text sound like a competent human peer: clear, concise, and grounded in reality.

Core Mandates

1. Authenticity over Polish: Prefer natural phrasing over corporate or marketing speak. Avoid passive voice when possible.
2. Eradicate Grandiosity: Remove words like "revolutionary", "shines", "unparalleled", "game-changer", "magic", "tapestry", "seamlessly", "supercharge", "unleash", or "delve". Replace them with specific, evidence-based descriptions.
3. Kill the Filler: State facts directly. Remove transitional fluff like "In conclusion", "Furthermore", "Surprisingly", "Interestingly", "It's important to note that", and "Additionally".
4. No Minimizing Language: Remove patronizing words like "simply", "just", "obviously", and "easy". State the action without judging its difficulty.
5. Fix Metaphors: Eliminate clichéd metaphors like "silver bullet", "missing link", or "piece of the puzzle". Describe the actual connection or realization instead.

Workflow

When asked to review or improve a text: