video-keyframe-analyzer

Extract representative keyframes from videos using OpenCV and Pillow.

3|Updated Sep 29, 2025

Search Tags:#video #whisper #opencv #analyze #keyframes #image-processing #cost-reduction

Ready to run

npx skills add https://github.com/xtone/ai_development_tools --skill video-keyframe-analyzer

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: video-keyframe-analyzer
Source: https://github.com/xtone/ai_development_tools/tree/main/common_development/skills/video-keyframe-analyzer
Command: npx skills add https://github.com/xtone/ai_development_tools --skill video-keyframe-analyzer

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill reduces unnecessary data and cost by automatically extracting only meaningful keyframes from videos for Claude Code analysis, enabling faster insights at a fraction of the data transfer cost.

Core Features & Use Cases

Keyframe extraction: Detect and save representative frames using frame-difference analysis.
Quality optimization: Downscale and compress frames to balance clarity and size.
Use Case: QA teams can validate UI flows or design reviews by comparing keyframes instead of full videos.

Quick Start

Run the script to generate keyframes: python extract_keyframes.py <video_path> --mode distributed
Or use timestamps to extract frames at specific moments: python extract_keyframes.py <video_path> --timestamps 0.0,1.5,3.0

Frequently Asked Questions about video-keyframe-analyzer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I extract keyframes from a video to reduce API costs?▼

Extract keyframes by running the Python script with distributed mode to detect representative frames using frame-difference analysis, which reduces data sent to Claude API and cuts costs.

What is the best way to analyze UI flows from screen-capture videos without processing every frame?▼

Extract keyframes from screen-capture videos using frame-difference analysis to isolate representative UI frames for QA verification, skipping redundant frames to minimize data transfer and API costs.

Can I extract video frames at specific timestamps instead of automatic detection?▼

Yes, you can extract frames at specific timestamps by passing a comma-separated list to the script. This targets exact moments for analysis instead of relying on distributed sampling or frame-difference detection.

Does this keyframe extraction tool support audio transcription for context?▼

Yes, it supports optional Whisper transcription to provide audio context alongside extracted keyframes, adding spoken context to visual analysis for screen-capture QA and UI flow verification.

Do I need Python and OpenCV installed to extract keyframes from videos?▼

Yes, you need Python with OpenCV, Pillow, and NumPy installed, as the script relies on these image-processing libraries for frame-difference analysis, downscaling, and compression.

Why should I extract keyframes before sending videos to Claude for analysis?▼

Extracting keyframes reduces unnecessary data and cost by sending only representative frames to Claude API instead of full videos, enabling faster insights at a fraction of the data transfer cost for QA reviews.