This skill integrates the VideoDB SDK into Claude Code, allowing developers to build video-native AI workflows through natural language. It provides a comprehensive suite of tools for ingesting media from local files, URLs, or RTSP feeds, and building semantic, visual, and spoken indexes for precise timestamped searches. Beyond analysis, the skill enables programmatic timeline editing—including subtitle generation, branding overlays, and transcoding—as well as real-time desktop perception for capturing and summarizing screen sessions. It is an essential toolkit for building AI agents that need to see, remember, and act upon video content.
주요 기능
01Semantic and visual indexing for timestamp-level video search
02Programmatic timeline editing for subtitles, overlays, and transitions
03Automated transcoding and reframing for social media platforms
04151 GitHub stars
05Multi-source ingestion from local files, YouTube, URLs, and RTSP feeds
06Real-time desktop capture and session memory (macOS)
사용 사례
01Automating video post-production tasks like subtitling and branding
02Building AI-driven monitoring systems for live security or broadcast feeds
03Generating searchable timelines and summaries from recorded meetings or desktop sessions