SproutVideo MCP Server

What is SproutVideo MCP Server?

A Model Context Protocol (MCP) server that wraps the SproutVideo API, enabling AI models to interact with SproutVideo content through standardized tools. It adds database persistence, semantic search, and security features, making it suitable for developers and AI systems that need programmatic access to video metadata, search, and management.

How to use SproutVideo MCP Server?

Clone the repository, install dependencies (npm install), set up PostgreSQL with the pgvector extension, create a .env file with your configuration, run database migrations (npm run migrate), build the project (npm run build), and start the server (npm start). The server listens for MCP requests on standard input/output channels. Use npm run sync to synchronize video metadata from SproutVideo to the local database.

Key features of SproutVideo MCP Server

• MCP‑compliant for seamless AI integration
• Tool‑based architecture with video and search tools
• Metadata persistence in PostgreSQL with pgvector
• Semantic search using vector embeddings
• High‑availability embedding system with automatic fallback
• Security layer: API key management, access control, audit logging

Use cases of SproutVideo MCP Server

• Retrieve detailed information about a specific SproutVideo video
• List and filter videos with pagination, ordering, folder, or tag filters
• Perform natural‑language semantic searches across video content
• Update video metadata (title, description, tags, privacy)
• Generate concise summaries of video content

FAQ from SproutVideo MCP Server

What tools does SproutVideo MCP Server provide?

It provides get_a_video, list_videos, search_videos, edit_video_metadata, and generate_video_summary.

What are the prerequisites for running the server?

Node.js v16+, npm or yarn, PostgreSQL with the pgvector extension, a SproutVideo API key, and either an Ollama instance or an OpenAI API key for embeddings.

How does semantic search work?

The server generates vector embeddings of video content using either Ollama or OpenAI, then performs similarity search to return results matching a natural language query.

Where is data stored?

Video metadata is persisted in a PostgreSQL database, and vector embeddings for semantic search are also stored there. Original video files remain on SproutVideo.

What security features are included?

The server implements API key management, access control, and audit logging to protect sensitive operations and track usage.