Create typographic self-portrait films for AI narratives, attention visualizations, and kinetic typography. A full-frame monospaced corpus wall → a query splits into word chips, growing curved attention lines toward other wall words → an answer types word by word beside a candidate probability list → the answer enters a highlighted center bubble, surrounded by a wall of chat bubbles that continuously shrinks to tile the frame → bubbles disperse radially → a breathing cursor in darkness → the corpus wall returns faintly for a typed closing. One continuous take with no hard cuts, a warm near-black background + grain + a single amber accent color, and a beatless dark noise music bed. Includes a render-ready Remotion scaffold with frame-by-frame canvas drawing, driven entirely by slot JSON. Triggers include text walls, corpus walls, attention visualization, attention lines, token probability lists, chat bubble walls, AI self-portrait films, kinetic typography films, and 30-second typographic films.

Język

• angielski

Tekst na ekranie: dowolny język

Technologia

• Remotion
• Google Fonts

Wymaga

• Bun
• Python 3
• FFmpeg

Wymaga

• Bez kluczy API

Działa w

• Claude Code
• Codex

Wynik

• 16:9
• MP4

Działa offline

• Nie

Downloads Google Fonts at render time

Testy

• Testy statyczne zaliczone · bez próbnego renderowania

Generuje za pomocą

• Your agent's own model