AI agents are already fluent in one thing: writing HTML. HyperFrames, HeyGen’s open-source framework, turns that skill into video — an agent writes an HTML file, HyperFrames renders a deterministic MP4. It’s been surging on GitHub this week as teams wire it into agent pipelines.
## How it works
A composition is just an HTML file with data attributes — no React, no proprietary timeline format. Under the hood it uses headless Puppeteer to capture frames and FFmpeg to encode, so the same HTML file always produces the same video. You bring your own animation runtime: GSAP, CSS animations, Lottie, Three.js, Anime.js, or WAAPI all plug in through adapters.
## Built for agents
The CLI is non-interactive by default, and HyperFrames ships skills that teach an agent the full production loop: plan the video, write valid HTML, wire seekable animations, add media, lint, preview, and render. It’s Apache 2.0 with no per-render fees or commercial thresholds, already running in production at HeyGen with community examples from teams like tldraw and TanStack. For agents that can code but couldn’t make video, that’s the missing primitive.

Leave a comment