video-recorder

name	video-recorder
description	Generate professional videos with branded slides and AI voiceover. Renders HTML presentations to video with synchronized audio using ElevenLabs TTS.
version	1.0.0
type	python
python_module	superskills.video_recorder.src:VideoRecorder
license	MIT

Video Recorder

Profile: Review PROFILE.md for brand customization.
Master Briefing: Global brand at ~/.superskills/master-briefing.yaml applies automatically.

Generate complete videos with branded slide presentations and professional voiceover.

Purpose

Enable autonomous video generation for training materials, workshop recaps, educational content, and client explainer videos. Combines branded visual slides with AI-generated ElevenLabs voice for polished, professional output.

Tools

VideoRecorder.py - Main orchestrator
SlideRenderer.py - HTML slides to PNG frames (Playwright)
AudioGenerator.py - Script to voiceover (ElevenLabs via narrator skill)
VideoEncoder.py - Frame + audio assembly (FFmpeg)
TimingSync.py - Slide-audio synchronization

Core Workflow

1. Generate Audio - Convert script to voiceover via ElevenLabs
2. Render Slides - HTML templates → PNG frames (branded)
3. Calculate Timing - Distribute slides across audio duration
4. Encode Video - Assemble frames + audio → MP4

Usage

Basic Usage

superskills call video-recorder '{
  "script": "Welcome to this presentation. Today we cover three key topics.",
  "slides": [
    {"type": "title", "heading": "Welcome"},
    {"type": "content", "heading": "Topics", "bullets": ["Planning", "Execution", "Results"]}
  ],
  "output_name": "my_video"
}'

Slide Types

Title Slide:

{
  "type": "title",
  "heading": "Welcome",
  "subheading": "Optional subtitle"
}

Content Slide:

{
  "type": "content",
  "heading": "Key Points",
  "bullets": ["Point 1", "Point 2", "Point 3"],
  "text": "Optional additional context"
}

Image Slide:

{
  "type": "image",
  "heading": "Diagram",
  "image_url": "/path/to/image.png",
  "caption": "Optional caption text"
}

Output

Returns JSON with:

• video_path: Full path to generated MP4
• duration_seconds: Total video duration
• slide_count: Number of slides
• audio_path: Path to voiceover audio
• metadata: Resolution, FPS, voice info

Quality Standards

• Resolution: 1920x1080 (Full HD)
• Format: MP4 (H.264 video, AAC audio)
• Audio: 192kbps, clear voiceover
• Branding: Consistent colors, fonts, your brand signature (configured in brand/default.yaml)
• Timing: Slides distributed evenly across audio

Examples

Single Slide Test

superskills call video-recorder '{
  "script": "This is a test of the video recording system.",
  "slides": [{"type": "title", "heading": "Test Video"}],
  "output_name": "test"
}'

Multi-Slide Presentation

superskills call video-recorder '{
  "script": "Welcome everyone. First, we discuss planning. Then implementation. Finally, results.",
  "slides": [
    {"type": "title", "heading": "Welcome", "subheading": "AI Video Generation"},
    {"type": "content", "heading": "Agenda", "bullets": ["Planning", "Implementation", "Results"]},
    {"type": "title", "heading": "Thank You"}
  ],
  "output_name": "presentation"
}'

Dependencies

• System: FFmpeg (install via brew install ffmpeg)
• Python: playwright, ffmpeg-python, pydub, jinja2
• Browser: Chromium (install via playwright install chromium)
• Voice: ElevenLabs API key (via narrator skill)

Limitations

• Static slides (no animations in MVP)
• Equal timing distribution (advanced timing in future)
• macOS/Linux/Windows compatible (Playwright cross-platform)
• Requires internet for voice generation

Troubleshooting

FFmpeg not found:

• Install: brew install ffmpeg
• Verify: ffmpeg -version

Playwright browser missing:

• Install: playwright install chromium

Voice generation fails:

• Check narrator skill configuration
• Verify ElevenLabs API key in ~/.superskills/.env

Video encoding fails:

• Check FFmpeg logs in error output
• Verify temp files in .{output_name}_work/ directory
• Ensure sufficient disk space

Future Enhancements

• Live window capture mode (using vision skill)
• Advanced timing with silence detection
• Slide transitions (fade, slide)
• Multiple voice support
• Background music
• CSS animations
• Auto-generated thumbnails
• Subtitle generation (SRT/VTT)