fluff-detector

Verify that agents, commands, and skills are written for LLM consumption, not human readers. Every token should improve execution, not provide comfort or credibility.

Content that establishes credibility for humans but has no LLM value.

| "Inspired by X" | "Inspired by DHH's approach" | LLM can't retrieve by author | | "Based on X's methodology" | "Based on Martin Fowler's patterns" | Pure credibility signal | | "According to X" | "According to industry experts" | Appeal to authority | | "As recommended by" | "As recommended by Google" | No execution value |

اكتشاف المحتوى الموجه نحو الإنسان في التحف LLM. إسناد الأعلام، والشخصيات، والتفسيرات الزائدة عن الحاجة، ولغة التسويق، والعناصر الزخرفية التي تهدر الرموز المميزة دون تحسين سلوك LLM. المصدر: majesticlabs-dev/majestic-marketplace.