Multimedia Processing Toolkit

What It Solves

Daily office and creator workflows often scatter images, slides, audio, and video across separate tools. This skill consolidates repeated tasks—batch conversion, resizing, watermarking, transcription, and styling—into scripts and reference docs, so you can validate a small sample before running a larger batch.

How It Works

The skill is script-first and task-oriented:
- Images: format conversion, compression, cropping, watermarking, and background replacement via flags such as --convert, --compress, --resize, and --watermark.
- Slides: check reports issues, beautify restyles a deck, palette suggests color schemes, and fontcheck inspects fonts.
- Audio/Video: transcribe creates transcripts, subtitle generates captions, speed changes pace, convert changes formats, and trim/concat edit clips.
- AI-assisted design: --prompt-zh or --prompt produces image-generation prompts and can support cover layout.

A typical run starts with --example to inspect usage, then executes the relevant script, followed by checking output format, image quality, caption timing, and cleanup.

Boundaries

This is best for office decks, social-media graphics, meeting notes, and light video editing. It is not a substitute for print production, cinematic editing, or commercial retouching. Background replacement works best on solid backgrounds; transcription and captioning may need network access, with --engine vosk as an offline option. Video editing may require ffmpeg or moviepy; image batches are safer under 100 files; slide beautify writes a new file instead of overwriting the source.