Embedding Search
Enables searching transcript segments using vector similarity via a Model Context Protocol (MCP) server.
About
This tool is an MCP server designed to query a Turso database containing embeddings and transcript segments. By leveraging vector similarity search, it allows users to find relevant transcript segments simply by asking questions, without requiring the generation of new embeddings. It provides complete transcript metadata, configurable search parameters, efficient database connection pooling, and comprehensive error handling, ensuring quick and accurate results.
Key Features
- Configurable search parameters (limit, minimum score)
- Relevance scoring based on cosine similarity
- Efficient database connection pooling
- Vector similarity search for transcript segments
- Complete transcript metadata (episode title, timestamps)
- 0 GitHub stars
Sponsored