| name | academy-video-pipeline |
| description | Programmatic HTML→video generation for the Zouroboros Academy and Alaric/Mimir explainer content, backed by HeyGen HyperFrames. Author a composition as a single HTML file (GSAP/anime.js timelines, captions, voiceovers, audio-reactive visuals) and render it to a real MP4/WebM/MOV/GIF locally with headless Chrome + ffmpeg — no manual ffmpeg zoompan pipeline, no cloud render. Use this to build title cards, faceless explainers, slideshows, product/launch videos, motion graphics, and captioned clips. The heavy multi-step orchestration skills (faceless-explainer, slideshow, etc.) live in Integrations/hyperframes/skills/. |
| compatibility | Created for Zo Computer |
| metadata | {"author":"marlandoj.zo.computer"} |
| allowed-tools | Bash, Read, Write, Edit |
Academy Video Pipeline (HyperFrames)
Replaces the hand-rolled ffmpeg-zoompan + ElevenLabs stills pipeline for
Academy / Alaric / Mimir explainer content. You write HTML, HyperFrames
renders video. Apache-2.0, runs fully local.
Validated environment (2026-06-24)
Confirmed working on this Zo host with no Docker:
- Node +
npx hyperframes@latest (pinned via npx, no global install needed)
- Headless Chrome auto-detected (chrome-headless-shell);
ffmpeg/ffprobe present
- A 3s 1080p30 composition renders in ~10s wall at
-q draft, 3 workers
doctor reports Docker + whisper-cpp ✗ — not needed for local render
(whisper-cpp only powers transcribe; TTS uses Kokoro-82M)
Always export PUPPETEER_SKIP_DOWNLOAD=1 so npx doesn't try to fetch its own
Chromium (the host's is used instead).
Quick render (the validated path)
export PUPPETEER_SKIP_DOWNLOAD=1
npx -y hyperframes@latest init my-video --non-interactive --example blank --skip-skills
cd my-video
npx -y hyperframes@latest lint
npx -y hyperframes@latest render -q draft -o renders/test.mp4
Output is a real H.264 MP4 (verify with ffprobe). Other formats:
--format webm|mov|gif|png-sequence. Tune with -f/--fps, -w/--workers,
-q/--quality.
When to use which orchestration skill
For anything beyond a single title card, drive one of the vendor multi-step
skills in Integrations/hyperframes/skills/ (they cross-reference each other
with relative paths, so run them from there). The Academy/Mimir fit:
- faceless-explainer — text → narrated faceless explainer (30-90s sweet
spot, every visual invented). Primary Academy/Mimir skill. SkillSpector
pre-adoption scan: score 17 / LOW / SAFE.
- slideshow — deck-style sequences. SkillSpector: 5 / LOW / SAFE.
- Also shipped: general-video, embedded-captions, graphic-overlays,
motion-graphics, music-to-video, product-launch-video, website-to-video,
pr-to-video, hyperframes-animation/creative/media/cli.
If the route is unclear, read Integrations/hyperframes/skills/hyperframes/
first — it's the router.
Security posture
HyperFrames is a Claude Code plugin marketplace (.claude-plugin/). The two
skills we lead with both pass the skill-security-gate (faceless-explainer 17,
slideshow 5 — both SAFE). Re-gate any other vendor skill before first use:
bash Skills/skill-security-gate/scripts/gate.sh Integrations/hyperframes/skills/<name>
Underlying tool
- CLI:
npx -y hyperframes@latest <cmd> (init, lint, render, preview, doctor,
inspect, transcribe, tts, capture, publish).
- Source clone:
Integrations/hyperframes/ (heygen-com/hyperframes, Apache-2.0).
- Docs: hyperframes.heygen.com
