Skip to main content

emotion-curve-feedback

Emotion Curve Feedback — Multi-Persona Emotional Journey Diagnostics for any sequential content. Simulate 2-4 audience personas advancing through the content unit by unit (slide, chapter/section, scroll block), score each unit's emotion on a -5..+5 scale per persona, render the smooth overlaid emotion curves with a matplotlib script, then read the valleys (where the content loses people), single-persona dips (persona conflict), and the landing height (how energized each persona is at the close). Turns those into priority-ordered fixes. Works on slide decks, articles, papers, landing pages, recruitment/donation pages, proposals, long announcements. Use when you want to evaluate content by the AUDIENCE's lived experience rather than the author's intent, or to find where it goes cold. This is the diagnostic counterpart to narrative-heat-engineering (NHE). Triggers: 「感情曲線で診断して」「どのスライド/セクションで冷めるか見て」「ペルソナ別に評価して」「谷を特定して」.

Jump to install

Source facts

Repository
NyxFoundation/skills
Last source activity
July 25, 2026 at 12:01
Detected SKILL.md language
Japanese
Stars
0
Forks
0

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.