Skip to main content
The Video Creator skill pack enables programmatic video creation and processing workflows. It combines media processing tools with object storage for complete video production pipelines.

​
Included Services

FFmpeg

Media transcoding, editing, and processing

Remotion

Programmatic video rendering with React

MinIO

S3-compatible object storage for media files

​
Skills Provided

​
FFmpeg Process

Capabilities:
  • Transcode video between formats (MP4, WebM, AVI, etc.)
  • Extract frames or frame sequences
  • Merge audio and video tracks
  • Extract audio from video
  • Trim and split video files
  • Get detailed media information
Example Usage: 
# Convert video to MP4 with H.264
ffmpeg -i /data/input/video.avi \
  -c:v libx264 -preset medium -crf 23 \
  -c:a aac -b:a 128k \
  /data/output/video.mp4

# Extract frames at 1 FPS
ffmpeg -i /data/input/video.mp4 \
  -vf "fps=1" \
  /data/output/frames/frame_%04d.png

# Extract audio as WAV for Whisper
ffmpeg -i /data/input/video.mp4 \
  -vn -c:a pcm_s16le -ar 16000 -ac 1 \
  /data/output/audio.wav

​
Remotion Render

Capabilities:
  • Define video compositions in React
  • Render programmatic videos to MP4/WebM
  • Create data-driven videos from JSON
  • Generate video thumbnails
  • Orchestrate multi-scene compositions
  • Add audio tracks and effects
Example Usage: 
// Define a composition
import { AbsoluteFill, useCurrentFrame, interpolate } from "remotion";

export const MyVideo: React.FC = () => {
  const frame = useCurrentFrame();
  const opacity = interpolate(frame, [0, 30], [0, 1]);

  return (
    <AbsoluteFill style={{ backgroundColor: "#0a0a0a" }}>
      <h1 style={{ color: "white", opacity }}>
        Hello from OpenClaw
      </h1>
    </AbsoluteFill>
  );
};

# Render to MP4
npx remotion render src/index.ts MyVideo /data/output/video.mp4

# Render with custom data
npx remotion render src/index.ts DataVideo /data/output/data-video.mp4 \
  --props='{"title":"Q4 Results","value":42}'

​
MinIO Storage

Capabilities:
  • Store and retrieve media files
  • S3-compatible API
  • Bucket management
  • Pre-signed URLs for temporary access
  • Multi-part upload for large files
Example Usage: 
# Upload a rendered video
mc cp /data/output/video.mp4 minio/openclaw-videos/

# Generate a shareable link (7 days)
mc share download minio/openclaw-videos/video.mp4 --expire=7d

# List all videos in bucket
mc ls minio/openclaw-videos/

​
Use Cases

​
Automated Video Production

Combine Remotion with data sources to create automated video content:
  1. Data Visualization Videos - Convert CSV/JSON data into animated charts
  2. Social Media Content - Generate branded video clips from templates
  3. Tutorial Videos - Programmatically create educational content
  4. Marketing Videos - Dynamic video generation from product data

​
Media Processing Pipeline

Build end-to-end media workflows:
  1. Upload raw video to MinIO
  2. Extract audio with FFmpeg
  3. Transcribe with Whisper (from Local AI pack)
  4. Generate subtitled video
  5. Store final output in MinIO

​
Video Thumbnails

Generate preview images and thumbnail sheets: 
# Extract single frame as thumbnail
ffmpeg -i /data/input/video.mp4 -ss 00:00:05 -frames:v 1 /data/output/thumb.jpg

# Create 4x4 thumbnail grid
ffmpeg -i /data/input/video.mp4 \
  -vf "select=not(mod(n\,100)),scale=320:180,tile=4x4" \
  -frames:v 1 /data/output/thumbnails.jpg

​
Example Workflow

Here’s a complete video processing workflow: 
# 1. Download source video from MinIO
mc cp minio/raw-footage/source.mp4 /data/input/

# 2. Extract audio for transcription
ffmpeg -i /data/input/source.mp4 \
  -vn -ar 16000 -ac 1 \
  /data/temp/audio.wav

# 3. Transcode to multiple formats
ffmpeg -i /data/input/source.mp4 \
  -c:v libx264 -crf 23 \
  /data/output/video-hq.mp4

ffmpeg -i /data/input/source.mp4 \
  -c:v libvpx-vp9 -crf 30 \
  /data/output/video.webm

# 4. Generate thumbnail
ffmpeg -i /data/input/source.mp4 \
  -ss 00:00:10 -frames:v 1 \
  /data/output/thumbnail.jpg

# 5. Upload all outputs to MinIO
mc cp /data/output/video-hq.mp4 minio/processed/
mc cp /data/output/video.webm minio/processed/
mc cp /data/output/thumbnail.jpg minio/thumbnails/

​
Configuration

​
Environment Variables

These variables are automatically configured in your generated stack: 
# FFmpeg (CLI tool, no host/port needed)
SHARED_VOLUME=/data

# Remotion (CLI tool, no host/port needed)
REMOTION_OUTPUT_DIR=/data/output/videos

# MinIO
MINIO_HOST=minio
MINIO_PORT=9000
MINIO_ROOT_USER=minioadmin
MINIO_ROOT_PASSWORD=<generated>

​
Volume Mounts

All services share a common volume for data exchange: 
volumes:
  openclaw_data:
    driver: local

​
Memory Requirements

  • FFmpeg: ~512 MB (more for 4K video)
  • Remotion: ~1 GB during render
  • MinIO: ~512 MB base + storage
Total: ~2 GB minimum

​
Performance Tips

​
FFmpeg

  • Use -c copy for operations that don’t need re-encoding (trimming, merging)
  • Set -preset fast or -preset veryfast for faster encoding
  • Use hardware acceleration if available: -hwaccel cuda

​
Remotion

  • Set concurrency in render options to use multiple CPU cores
  • Render still frames instead of full videos for thumbnails
  • Use <Sequence> to break long videos into manageable segments

​
MinIO

  • Use multi-part upload for files larger than 100 MB
  • Enable versioning for important media assets
  • Set lifecycle policies to auto-delete old files

​
Next Steps

Local AI Pack

Add Whisper for video transcription

Services

Explore all available services

Build docs developers (and LLMs) love

Get started for freeTalk to us