from VideoDB72
Agent skill for ingesting, indexing, searching and programmatically editing video and desktop capture; returns playable streams, timestamps, clips, and transcod
VideoDB provides perception, indexing, and action capabilities for video, live streams (RTSP), and desktop capture. It can ingest files or streams, transcode/normalize media, extract spoken transcripts and visual scene indexes, search for moments with timestamps and playable evidence, and compose timeline edits (subtitles, overlays, dubbing) programmatically.
Use VideoDB when agents must understand video content, produce searchable timelines, generate clips, or perform server-side edits and transcodes. Typical scenarios: monitoring live feeds with alerts, producing session summaries from desktop recordings, automated subtitle generation and burn-in, and reformatting videos for social platforms.
scripts/ helpers and a WebSocket listener for desktop capture are present.reference/ directory documents indexing, editor workflows, streaming and capture.Designed for integration with agent runtimes that can run Python or call APIs; compatible with Claude/Codex/Gemini-backed agents or any agent that can call the VideoDB Python SDK or HTTP API.
This skill has not been reviewed by our automated audit pipeline yet.
