prest

Why we included this project

pREST sits in front of a PostgreSQL database and exposes your tables, views, and custom SQL routes as HTTP endpoints, so you skip hand-writing the CRUD layer. Auth and access control are built in, and because it runs as a Go binary or Docker container, it drops into an existing stack without forcing a new framework. It also generates an OpenAPI schema, which front-end and mobile teams can build against directly. For AI work, the same server offers a Model Context Protocol endpoint that lets Claude, Cursor, and other MCP-capable clients discover schemas and read rows through the same auth and ACL rules you already apply to REST. That makes it both a quick backend for internal tools and a governed read bridge between AI assistants and the data they need.