Skills website-to-hyperframes
🎬

website-to-hyperframes

Low Risk ⚙️ External commands🌐 Network access📁 Filesystem access

Create professional videos from any website

Converting website designs into engaging videos requires complex technical steps from capture to render. This skill automates the entire 7-step pipeline, producing professional videos with narration, animations, and brand-consistent styling from a simple URL.

Supports: Claude Codex Code(CC)
🥉 73 Bronze
1

Download the skill ZIP

2

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

3

Toggle on and start using

Test it

Using "website-to-hyperframes". Capture https://stripe.com and make me a 20-second product demo

Expected outcome:

A complete HyperFrames project with: capture/ folder (screenshots, assets, tokens), DESIGN.md, SCRIPT.md, STORYBOARD.md, compositions/ with animated HTML, narration.wav, transcript.json, and localhost Studio preview URL.

Using "website-to-hyperframes". Turn this website into a social ad: https://example.com

Expected outcome:

A 15-second Instagram Stories video (1080x1920) with brand-consistent animations, optional hook narration, and smooth transitions between beats.

Security Audit

Low Risk
v1 • 4/27/2026

Security evaluation confirms this skill is safe for publication. Static analysis flagged 409 potential issues, but all high-severity findings are false positives. The skill uses legitimate HyperFrames CLI tooling (npx hyperframes capture/lint/validate) for video production workflows. External commands are standard CLI tooling, network access is limited to website capture, and env_access is optional API key configuration. No malicious code execution, credential exfiltration, or data harvesting patterns detected.

9
Files scanned
1,500
Lines analyzed
5
findings
1
Total audits
Low Risk Issues (2)
External Command Patterns in Documentation
287 references to shell commands (npx hyperframes) flagged by static scanner. These are legitimate CLI tooling invocations documented for the video production workflow, not malicious code execution.
Optional API Key Environment Access
References to GEMINI_API_KEY in step-1-capture.md:19. This is an optional, user-provided enhancement for AI-powered image descriptions. Not hardcoded secrets.

Detected Patterns

False Positive: Weak Cryptographic Algorithm FlagsFalse Positive: Screen Capture Upload FlagsFalse Positive: Ruby/Shell Backtick Execution
Audited by: claude

Quality Score

45
Architecture
100
Maintainability
87
Content
50
Community
86
Security
91
Spec Compliance

What You Can Build

Product launch videos

Capture a product website and create a polished launch video with animations, logo reveals, and product feature highlights.

Social media ads

Transform any website into Instagram or TikTok ads with kinetic typography, brand colors, and optional voiceover.

Website portfolio videos

Create video walkthroughs of portfolio sites for use in presentations, pitches, or social proof.

Try These Prompts

Simple website capture
Capture https://example.com and make me a 25-second product launch video
Social ad with specific format
Turn this website into a 15-second social ad for Instagram: https://mysite.com
Product demo with voiceover
Create a 30-second product demo video from https://product.com using the default voiceover voice
Full production with custom timing
Capture https://landingpage.io and produce a 45-second brand reel with: 3 intro beats, 2 feature highlights, 1 CTA. Use portrait format 1080x1920.

Best Practices

  • Start with websites that have clear visual hierarchy and distinct sections for best capture results
  • Review captured assets and design tokens before writing the script to ensure brand consistency
  • Use the validation step (Step 7) to catch layout issues before rendering to MP4

Avoid

  • Do not skip the capture step and provide raw URLs to sub-agents - always capture first
  • Do not inline assets (SVG/image data) in compositions - always reference captured files by path
  • Do not use Google Fonts when captured fonts are available - use @font-face with local woff2 files

Frequently Asked Questions

What websites work best with this skill?
Static marketing sites, landing pages, and product pages work best. Sites with heavy JavaScript interactions or login walls may have limited capture quality.
Do I need any API keys?
No API keys are required for basic capture and video production. Optionally, you can set GEMINI_API_KEY for AI-powered image descriptions (~$0.001/image).
What video formats are supported?
Landscape (1920x1080), Portrait (1080x1920 for Instagram Stories/TikTok), and Square (1080x1080 for Instagram feed).
Can I add my own narration?
Yes. The skill generates TTS, but you can replace narration.wav with your own voiceover recording. Update the transcript.json with your word-level timestamps.
How long does rendering take?
Capture takes 10-30 seconds. Building compositions takes 2-5 minutes depending on complexity. Final MP4 rendering via Docker takes 1-3 minutes.
What's the difference between the preview URL and the MP4 file?
The localhost Studio URL (http://localhost:port) is the live preview - share this first for review. The MP4 file is the final deliverable - render on explicit request only.