How do I authenticate the YouTube skill?

You must install the atris CLI and run 'atris login' to generate the necessary credentials before using the skill within Claude Code.

How are credits handled for video processing?

Each video processing task costs 5 credits. If the processing fails or the API returns an error, the credits are automatically refunded to your account.

Does this skill only read video transcripts?

No, it uses Gemini's native multimodal API to process actual video frames and audio, allowing it to understand visual demonstrations, slides, and non-verbal context.

What happens if a video is very long?

The skill charged a flat rate of 5 credits per video regardless of length, as long as it is supported by the underlying Gemini API.

Can I save video insights for future use?

Yes, by providing an agent_id and setting store_as_knowledge to true, you can save the analysis directly into your agent's persistent knowledge base.

YouTube Video Intelligence

Name: YouTube Video Intelligence
Author: atrislabs

byatrislabs

•

Aprendizaje y Documentación

Analyzes YouTube videos using multimodal AI to extract visual and audio insights for research and knowledge management.

The YouTube skill for Claude Code leverages Gemini's native multimodal API to process video content directly, moving beyond simple transcript scraping by analyzing both visual frames and audio. This allow Claude to accurately describe on-screen demonstrations, read presentation slides, and answer complex questions about video content with high precision. Integrated with the atris context system, it enables users to seamlessly convert video content into structured knowledge that can be stored in an agent's memory for future reference and automated workflows.

Características Principales

01Direct knowledge base storage for persistent agent memory

02Focused query support for specific video-based questions

03High-accuracy extraction of on-screen text and demonstrations

0458 GitHub stars

05Native multimodal video processing (visual and audio analysis)

06Automated credit refund system for failed processing attempts

Casos de Uso

01Converting technical video tutorials into searchable documentation

02Extracting key insights and data points from recorded webinars

03Building a research library from multiple topical YouTube videos

Características Principales

01Direct knowledge base storage for persistent agent memory

02Focused query support for specific video-based questions

03High-accuracy extraction of on-screen text and demonstrations

0458 GitHub stars

05Native multimodal video processing (visual and audio analysis)

06Automated credit refund system for failed processing attempts

Casos de Uso

01Converting technical video tutorials into searchable documentation

02Extracting key insights and data points from recorded webinars

03Building a research library from multiple topical YouTube videos