mrbeast-perspective

This Skill applies MrBeast's CTR framework: keep titles under 8 words with numbers first, ensure the thumbnail complements rather than repeats the title, and run the self-click test. The thumbnail_audit.py script scores title-thumbnail complementarity and curiosity gaps.

The Skill applies MrBeast's hook structure: show the core concept visually in 0-3 seconds, state the stakes by 8 seconds, flash the best moments by 15 seconds, and start action by 30 seconds. The retention_curve_checker.py script scores your script against these criteria.

The included analyze_titles.py script processes a text file of titles and reports length distribution, number usage, high-frequency words, formula classification (challenge, numeric, curiosity, contrast, emotional), and punctuation patterns as a Markdown report.

The Skill explicitly notes its limits: the formulas are optimized for YouTube, and platforms like Bilibili, Douyin, or TikTok have different algorithms and user behavior. Core principles like CTR × AVD transfer, but execution needs translation.

The Python scripts use only the standard library, except thumbnail_audit.py which optionally needs Pillow for image analysis. The subtitle fetcher shell script requires yt-dlp, which it installs automatically via pip if missing.