tzurot-db-vector

What problem does it solve?

Storing and retrieving long-term memory for AI personalities efficiently requires specialized database techniques like vector embeddings. This Skill provides patterns for PostgreSQL, Prisma, and pgvector, enabling scalable and intelligent memory retrieval.

Core Features & Use Cases

• Prisma ORM: Achieve type-safe database access for all data operations, reducing boilerplate and improving developer experience.
• pgvector Integration: Store and query AI embeddings for fast similarity search, enabling long-term, contextual memory for personalities.
• Connection Management: Implement robust connection pooling for Railway/containerized environments to prevent connection limits and cold start issues.
• Migration Workflow: Manage database schema changes with a reliable, checksum-safe Prisma migration process, ensuring database consistency across deployments.

Quick Start

Use the tzurot-db-vector skill to store a new AI memory embedding for a personality, then query for the top 5 most similar memories based on a given embedding.