Miguel Ángel
cb8e7995a2
* perf(runtime): swap edited scenes in place from a rebuilt preview document window.__hfSwapScenes(html) compares the preview manifests. When the rebuilt document differs only inside existing scenes, it replaces each changed scene's host, styles and scripts, drops the old scene timelines (nested ones included), reruns the passes boot runs on scene DOM, releases the old media, rebinds and keeps the playhead. Anything else rejects before touching the film, so the caller does its full reload. * fix(runtime): refuse unsafe scene swaps and redraw the swapped scene like a seek The swap now refuses scenes marked unswappable or with more than one host, keeps the style order, lets adapters find the new scene, redraws through the transport seek and rejects when torn down mid-swap. * fix(runtime): scope swap caption overrides to the swapped scene and swap only manifest pages Other scenes kept re-dimmed caption words after a swap, and pages without a scene manifest offered a swap that could only be refused. * fix(runtime): repaint caption colours that were overridden after their tween rendered A swapped scene, or a film reloaded mid-way, rendered its caption tweens before the overrides rewrote their colours, so words kept the old colour. * fix(runtime): repaint caption colours of a swapped scene the playhead has passed * refactor(core): cut scene swap comments to what names and types do not say * fix(runtime): a scene swap keeps other media loaded and waits for new media by reloading * fix(runtime): a scene swap resolves bound and proxied media before refusing The new-media check compared the new scene raw src against the live DOM, which variable bindings and the media proxy have already rewritten. A scene with a customised data-var-src logo, or a proxied video, was refused on every edit. Both sides now resolve src the same way: the bound variable value, else the src attribute before any proxy swap, which the proxy now records where it rewrites the src. * fix(runtime): a scene swap binds variables only in the swapped scene, before media setup The swap re-ran variable bindings over the whole document after binding media, so a proxied video in any scene got its original, undecodable src back and a bound text in an untouched scene was rewritten. Bindings now run only inside the swapped hosts and before media binding, as boot does, and the src guard compares against the src the element had before its proxy. * fix(runtime): a scene swap keeps unchanged video and audio playing A text edit beside a video rebuilt the video from scratch, so it went blank until it loaded again. A video or audio in the edited scene whose src matches one in the old scene (same src, same order among those) is now the old element, moved into the new scene and given the new element attributes and children, except src and the runtime-owned preload. * fix(runtime): a scene swap refuses more new-media forms and names the scene The new-media check missed uppercase URL(), image-set() and @import strings, SVG image href and xlink:href, and object data, so a new asset could pop in after the swap. Its refusal now says the media is new to this scene, which is what it checks. The existing inline style url(), poster and srcset checks get tests of their own. * fix(runtime): a scene swap loads caption overrides before changing the page The swap wrote the new scene first and only then fetched caption-overrides.json, so a stalled request left the preview half swapped when the caller gave up and reloaded. The overrides are now fetched before the first write and applied from the fetched data. A swap that another swap finished first while it waited now refuses, since its document is the older one. * refactor(runtime): keep the proxy src record within the comment share * fix(runtime): a scene swap keeps media only when the edit left it as written and ungraded Keeping a live video or audio and copying the new attributes onto it was not the same as a fresh load: muted only sets the initial state, and a graded video left its grading canvas in the discarded scene, so it went black. The swap now keeps an element only when its authored markup, captured before the runtime writes to it, equals the new one, and the grading runtime says it is not graded. Anything else is rebuilt fresh; the attribute copy is gone. * fix(runtime): a waiting scene swap refuses after any other swap changed the page The check compared the scene manifest, which misses a scene changed and changed back while a captioned swap waited for its overrides; that swap then went on with stale elements and wrote a wrong manifest. A generation count the swap owns now moves on every page change by a swap, so any intervening swap makes the waiting one refuse. * fix(runtime): a scene swap warns about a non-media data-var-src once, as a load does The new-media scan resolved every bound src through the warning path, so each swap logged the ignored data-var-src again for the old and the new element. The scan now resolves silently and only the bindings pass warns, once for the new element. * refactor(runtime): keep the grading query within the comment share * fix(runtime): a scene swap replaces each of a scene's styles in its own place The bundle can now give one scene several style runs separated by another scene's, when its nested scene comes later in source order. The swap put all new styles before the first old one, so a later same-named @keyframes moved ahead of the other scene's and the cascade differed from a fresh load and the render. Each new style now replaces its old counterpart in place, and a scene whose style count differs is refused before any write. * test(core): check the bundled stylesheets, not the inlined runtime, for a leftover @import * fix(runtime): a video kept through a scene swap starts its new tweens from its authored values * fix(runtime): a from() caption colour tween with an override still lights its word up * fix(runtime): a scene swap reverts the old timelines, so a kept video shows the edited CSS * fix(runtime): a caption word with a dim-colour override still lights up when spoken * fix(runtime): a word whose colour override lands late still looks dim before it is spoken * fix(runtime): a caption override keeps class highlights and resolves var() colours per word
Write HTML. Render video. Built for agents.
Quickstart | Showcase | Playground | Catalog | Docs | Discord
HyperFrames is an open-source framework for turning HTML, CSS, media, and seekable animations into deterministic MP4 videos. Use it locally with the CLI, from AI coding agents with skills, or as the rendering core behind hosted authoring workflows.
Quick Start
With an AI coding agent
Install the HyperFrames skills, then describe the video you want:
npx skills add heygen-com/hyperframes
The picker opens with nothing pre-selected — the Core Skills group is all you need: the
/hyperframesrouter installs each creation workflow on demand. Agents and non-interactive runs should usenpx hyperframes skills updateinstead — it installs exactly the core set, whereasskills add --allinstalls all 21 published skills. The six repo-internal skills under.claude/skills/.agents/skillsare excluded by default. For the full published set usenpx hyperframes skills.
skills addresolves the skills.sh registry blob, which can lagmainby hours.npx hyperframes skills updateinstalls from the currentmain, so reach for it when you need the newest copy of a skill.
Try a prompt like:
Using
/hyperframes, create a 10-second product intro with a fade-in title, a background video, and subtle background music.
The skills teach agents the HyperFrames production loop: plan the video, write valid HTML, wire seekable animations, add media, lint, preview, and render. They work with Claude Code, Codex, Cursor, Gemini CLI, IBM Bob, and other coding agents that support skills.
Skills
HyperFrames ships 21 skills agents load on demand. Read /hyperframes first — it's the router and capability map; it picks a workflow for any "make me a…" request — video, deck, or composition port — and points to the domain skills below.
Default to the core set — the router installs each creation workflow on demand. npx hyperframes skills update installs exactly that from anywhere; the interactive picker (npx skills add heygen-com/hyperframes) lists it as the "Core Skills" group, nothing pre-selected. The picker is interactive-only — a non-interactive or agent run without --skill installs all 21. Use npx skills add heygen-com/hyperframes --all to install all 20 deliberately (skips the picker), or npx skills add heygen-com/hyperframes --skill <name> for just one (bare name, no leading /).
Installs stay lean after that: npx hyperframes init keeps the core set fresh (the router, the hyperframes-* domain skills, and media-use — plus whatever is already installed; /figma stays on demand) and never expands a partial install; the creation workflows install on demand — the router runs npx hyperframes skills update <workflow> before entering one. Nothing re-pulls the full set behind your back.
Upload to Codex
Build the upload-ready Codex plugin archive from the committed HEAD version of the manifest, brand assets, and skills:
bun run package:codex-plugin
This writes dist/hyperframes-plugin.zip with a hyperframes/ root folder and fails if the archive exceeds Codex's 100 MB upload limit.
Router
| Skill | Use when |
|---|---|
/hyperframes |
Read first for any request to make / create / edit / animate / render a video, animation, or motion graphic. Capability map for the domain skills, the intent layer that confirms every creation brief up front, and intent router for the creation workflows below. |
Creation workflows
| Skill | Use when |
|---|---|
/product-launch-video |
Any website — marketing / launching / promoting a product (from its URL, a brief, or a script), or a site tour / showcase / social clip featuring the site's own visuals. Up to ~3 min (sweet spot 30-90s). |
/faceless-explainer |
Explaining a topic / concept from arbitrary text — no product, no URL, no website capture; every visual is LLM-invented (typography / abstract / diagram / data-viz). |
/pr-to-video |
A GitHub pull request (PR URL, owner/repo#N ref, or "this PR") → changelog / feature-reveal / fix / refactor explainer, read via the gh CLI. |
/embedded-captions |
Adding captions / subtitles to an existing talking-head video (footage untouched) — verbatim rail, embedded climax behind the subject, or pure-cinematic embed. |
/talking-head-recut |
Packaging an existing talking-head / interview / podcast video with designed graphic overlays — lower-thirds, data callouts, kinetic titles, pull-quotes, side panels, PiP. |
/motion-graphics |
A short, unnarrated, design-led motion graphic (~under 10s) — kinetic type, stat / chart hit, logo sting, lower-third, animated tweet / headline. MP4 or transparent overlay. |
/music-to-video |
A music track (audio file, video to pull audio from, or one generated from a mood brief) → a beat-synced video — lyric, slideshow, or kinetic promo; music drives pacing. |
/slideshow |
A presentation / pitch deck / interactive deck — discrete slides, fragment reveals, branching, hotspot navigation, presenter mode. Output is a navigable deck, not a rendered video. |
/general-video |
Anything else — longer or multi-scene pieces, brand / sizzle reel, title card, static loop, freeform composition. Input- and length-agnostic fallback, and the home of companion mode (co-create with the full toolbox). |
/remotion-to-hyperframes |
Porting an existing Remotion (React) composition's source to HyperFrames HTML. One-way migration, not creation. |
Domain skills (loaded on demand)
Atomic capabilities the creation workflows compose against — pull one when you need that specific layer.
| Skill | Covers |
|---|---|
/hyperframes-core |
The composition contract — data-* timing attributes, class="clip", tracks, sub-compositions, variables, framework-owned media playback, determinism rules. |
/hyperframes-animation |
All animation knowledge — atomic motion rules, scene blueprints, transitions, runtime adapters (GSAP / Lottie / Three.js / Anime.js / CSS / WAAPI / TypeGPU). |
/hyperframes-keyframes |
Seek-safe keyframe authoring across runtimes — GSAP timelines, CSS keyframes, Anime.js, WAAPI, FLIP, paths, masks, SVG morph/draw, 3D depth — plus hyperframes keyframes diagnostics for rendered motion. |
/hyperframes-creative |
Non-animation creative direction — frame.md / design.md, palettes, typography, narration, beat planning, audio-reactive visuals, composition patterns. |
/media-use |
The media OS — resolve any media need (BGM, SFX, image, icon, logo, voice, color grade, LUT) into a frozen local file or paste-ready block + ledger record, generate via TTS/music/image models when the catalog misses, transcribe, caption, remove backgrounds, and reuse assets across projects. One shared audio engine + manifest tracking. |
/hyperframes-cli |
CLI dev loop — init, lint, check, snapshot, preview, render, publish, doctor, plus HeyGen-hosted cloud rendering (cloud render) and AWS Lambda rendering (lambda deploy / render / progress). |
/hyperframes-audio |
Mix the audio already placed in a composition — voiceover carve (dip a music bed only in the bands the voice occupies, static or dynamic, level match included), the effect chain (EQ, compressor, limiter, gate, saturation, delay, reverb, chorus, phaser, bitcrush), automation envelopes on volume or any effect parameter, and submix buses (<hf-audio-group>) carrying one chain, fader and automation clock for several tracks at once. Sourcing the audio is /media-use. |
/hyperframes-registry |
Search, install and wire registry blocks and components into compositions via hyperframes catalog / hyperframes add. Load before hand-building any named look, effect, treatment or transition. Authoring a new block or component to contribute upstream. |
/figma |
Import Figma assets, tokens, components, and storyboard sections → reconstructed motion (frames read as states, not slides) (REST/CLI) plus Motion animations (MCP) and shaders (MCP source / native export) into a composition. |
For visual design handoff workflows, see the Claude Design guide and Open Design guide.
Manually with the CLI
npx hyperframes init my-video
cd my-video
npx hyperframes preview # preview in browser with live reload
npx hyperframes render # render to MP4
Requirements: Node.js 22+, FFmpeg
What You Can Build
Need ideas? Browse the Showcase for finished videos you can watch, read, run, and remix.
- Product launch videos and feature announcements
- PR walkthroughs with animated code diffs, narration, and captions
- Data visualizations, chart races, and map animations
- Social videos with kinetic captions, overlays, and music
- Docs-to-video, PDF-to-video, and site-tour explainers
- Reusable motion graphics for automated content pipelines
Frame.md
frame.md — your design system, ready for video.
Every brand has a design.md. None of them were written for a camera. frame.md is the missing translation layer: it takes your web-context design spec and inverts it for the frame — the same tokens, the same rules, but rewritten so an AI agent can compose a promo video without guessing at scale or reaching for web chrome.
The output is a DESIGN.md superset your whole toolchain can read. Atoms stay sacred. Composition stays free. Numbers come from the script.
Biennale Yellow |
BlockFrame |
Blue Professional |
Bold Poster |
Broadside |
Capsule |
Cartesian |
Cobalt Grid |
Coral |
Creative Mode |
Browse and remix them all at hyperframes.dev/design.
How It Works
Define a video as HTML. Add data attributes for timing and tracks. Use GSAP, CSS, Lottie, Three.js, Anime.js, WAAPI, or your own frame adapter for seekable animation.
<div id="stage" data-composition-id="launch" data-start="0" data-width="1920" data-height="1080">
<video
class="clip"
data-start="0"
data-duration="6"
data-track-index="0"
src="intro.mp4"
muted
playsinline
></video>
<h1 id="title" class="clip" data-start="1" data-duration="4" data-track-index="1">Launch day</h1>
<audio
data-start="0"
data-duration="6"
data-track-index="2"
data-volume="0.5"
src="music.wav"
></audio>
<script src="https://cdn.jsdelivr.net/npm/gsap@3/dist/gsap.min.js"></script>
<script>
const tl = gsap.timeline({ paused: true });
tl.from("#title", { opacity: 0, y: 40, duration: 0.8 }, 1);
window.__timelines = window.__timelines || {};
window.__timelines.launch = tl;
</script>
</div>
Preview instantly in the browser. Render locally or in Docker. The renderer seeks each frame in headless Chrome and encodes the result with FFmpeg, so the same input produces the same video.
HyperFrames Stack
HyperFrames is the open-source rendering engine, plus a growing set of tools around HTML-native video creation.
| Piece | Status | What it does |
|---|---|---|
| CLI | Available | Scaffold, preview, lint, inspect, and render local video projects |
| Core / Engine / Producer | Available | Parse compositions, drive headless Chrome, encode video, and mix audio |
| Catalog | Available | Reusable blocks and components for transitions, overlays, captions, charts, maps, and effects |
| Agent skills | Available | Teach coding agents the video-production patterns that generic web docs miss |
| Studio | Available, evolving | Browser surface for previewing and editing compositions |
| AWS Lambda rendering | Available | Deploy a distributed render stack and drive renders from your laptop or CI |
| hyperframes.dev | Available | Community playground for previewing, iterating, sharing, and rendering HTML-native video projects |
| frame.md | Available | Invert your design system for the camera — a DESIGN.md superset an agent can compose video from |
Catalog
Install ready-to-use blocks and components:
npx hyperframes add flash-through-white # shader transition
npx hyperframes add instagram-follow # social overlay
npx hyperframes add data-chart # animated chart
Browse the catalog at hyperframes.heygen.com/catalog.
Why HyperFrames?
- HTML-native: compositions are HTML files with data attributes. No React requirement, no proprietary timeline format.
- Agent-friendly: agents already write HTML, and the CLI is non-interactive by default.
- Deterministic: same input, same frames, same output. Built for CI, regression tests, and automated rendering.
- No build step: an
index.htmlcomposition plays as-is and can be previewed directly in the browser. - Adapter-based animation: bring GSAP, CSS animations, Lottie, Three.js, Anime.js, WAAPI, or a custom runtime.
- Open source: Apache 2.0 license, with no per-render fees or commercial-use thresholds.
HyperFrames vs Remotion
HyperFrames is inspired by Remotion. Both tools render video with headless Chrome and FFmpeg. The main difference is the authoring model: Remotion's bet is React components; HyperFrames' bet is plain HTML that humans and agents can both write easily.
| HyperFrames | Remotion | |
|---|---|---|
| Authoring | HTML + CSS + seekable animation | React components |
| Build step | None; index.html plays as-is |
Bundler required |
| Agent handoff | Plain HTML files | JSX / React project |
| Library-clock animations | Seekable, frame-accurate via adapters | Wall-clock animation patterns need care |
| Distributed rendering | Local and AWS Lambda render paths | Remotion Lambda, mature cloud renderer |
| License | Apache 2.0 | Source-available Remotion License |
Read the full comparison in the HyperFrames vs Remotion guide.
Documentation
Full documentation: hyperframes.heygen.com/introduction
Packages
| Package | Description |
|---|---|
hyperframes |
CLI for creating, previewing, linting, and rendering compositions |
@hyperframes/core |
Types, parsers, generators, linter, runtime, and frame adapters |
@hyperframes/engine |
Seekable page-to-video capture engine using Puppeteer and FFmpeg |
@hyperframes/producer |
Full rendering pipeline for capture, encode, and audio mix |
@hyperframes/studio |
Browser-based composition editor UI |
@hyperframes/player |
Embeddable <hyperframes-player> web component |
@hyperframes/shader-transitions |
WebGL shader transitions for compositions |
@hyperframes/aws-lambda |
AWS Lambda SDK and deployment surface for distributed renders |
Community
HyperFrames is used in production at HeyGen, with community examples from teams like tldraw, TanStack, and others in ADOPTERS.md. Open a PR if your team is using HyperFrames.
- Questions and ideas: Discord
- Bugs and feature requests: GitHub Issues
- User research: Book a casual 30-minute conversation with the HyperFrames team — no preparation or sales pitch
- Security reports: SECURITY.md
- Contributions: CONTRIBUTING.md
Development Note
The repo uses Git LFS for golden regression-test baselines under packages/producer/tests/**/output.mp4 (about 240 MB of .mp4 files). If you're cloning the full repo for development, install Git LFS first:
# macOS
brew install git-lfs
# Ubuntu / Debian
sudo apt install git-lfs
# Windows
winget install GitHub.GitLFS
# Then, once per machine
git lfs install
If you only need source files, you can skip LFS content:
GIT_LFS_SKIP_SMUDGE=1 git clone https://github.com/heygen-com/hyperframes.git