Extract frames and OCR text from MP4 videos

Video Content Extractor automatically processes MP4 video files by extracting key frames at configurable time intervals and performing OCR text recognition on each captured frame. It generates a comprehensive structured Markdown report that includes video metadata (duration, resolution, codecs) and frame-by-frame OCR transcripts with precise timestamp references.

Use this skill when you need to extract and index text content from video files, such as analyzing screen recordings, tutorial videos, presentation captures, or any video containing readable text. It's ideal for creating searchable transcripts of visual content, documenting software demonstrations, or archiving information displayed in video format.

Do not use this skill if your videos lack visible text content, as OCR will produce minimal useful output. This is not suitable for audio transcription - it only extracts visible text from video frames, not spoken words.

You provide MP4 video files and configure the time interval for frame extraction. The skill outputs a structured Markdown report containing video metadata (duration, resolution, codecs) and OCR-recognized text from each frame with corresponding timestamp references.

This skill is designed for Codex CLI and requires two local dependencies: FFmpeg for video frame extraction and processing, and Tesseract OCR for optical character recognition on the extracted frames. Both tools must be installed on your local machine before using this skill.

This skill serves developers, technical writers, QA engineers, and content creators who work with video content containing text. It's particularly valuable for teams documenting software interfaces, analyzing screen recordings, or building searchable archives of video tutorials and presentations where visual text content needs to be extracted and indexed.