videodb

Index, search, and orchestrate video assets via VideoDB collections.

16|3|Updated Mar 12, 2026

Search Tags:#streaming #transcription #websocket #videodb #video-indexing #live-capture #visual-indexing

Ready to run

npx skills add https://github.com/sehoon787/my-claude --skill videodb-sehoon787

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: videodb
Source: https://github.com/sehoon787/my-claude/tree/main/skills/ecc/videodb
Command: npx skills add https://github.com/sehoon787/my-claude --skill videodb-sehoon787

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires python-dotenv, videodb, and includes scripts (resource) and references (resource) components.

What problem does it solve?

VideoDB provides an integrated solution to make video content searchable, indexable, and automatically composable by indexing speech, visuals, and audio, enabling fast retrieval and generation of clips, streams, and summaries.

Core Features & Use Cases

Automatic transcription and spoken-word indexing for semantic search and keyword queries.
Visual and audio indexing to create scene-level descriptions and searchable metadata.
Timeline editing, subtitle generation, and on-demand streaming to assemble and share media.
Use Case: Build a searchable video library from hours of recordings and generate curated highlights with one command.

Quick Start

Install the VideoDB skill and connect to your collection, then start a capture or indexing workflow.

Frequently Asked Questions about videodb

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

FAQPage Schema

How do I make video content searchable by spoken words and visuals?▼

Video indexing makes video content searchable by automatically transcribing speech and generating scene-level visual descriptions for fast semantic retrieval. You can index both stored recordings and live capture streams to build a fully searchable video library.

Can I transcribe and index live capture streams in real time?▼

Yes, live capture streams can be transcribed and indexed in real time using websocket connections. The Skill connects to VideoDB to orchestrate live media ingestion alongside stored video assets for immediate spoken-word indexing.

How do I generate curated video highlights from hours of recordings?▼

You can generate curated highlights by using timeline-based rendering to assemble indexed video segments. Once speech and visual indexing is complete, retrieve specific scenes and compile them into a shareable clip with a single command.

Do I need a VideoDB collection to ingest and search video assets?▼

Yes, you need access to an existing VideoDB collection to perform video ingestion and search. The Skill also requires installing the videodb and python-dotenv Python packages to authenticate and connect to your collection.

What is the best way to add subtitles and metadata to a searchable video library?▼

Automatic transcription provides the spoken-word text needed for subtitle generation and searchable metadata. By indexing speech, visuals, and audio, the system creates scene-level descriptions that serve as both metadata and subtitle tracks.

Can I use this for on-demand streaming of assembled video clips?▼

Yes, on-demand streaming is supported for assembled media. After using timeline editing to compile indexed scenes into a new video, you can stream the resulting clip directly to share the curated content.