data-api-builder

Why we included this project

Surfacing an existing database as a usable API usually means standing up a custom backend. Data API builder skips that: point it at a configuration file and it exposes tables, views, and stored procedures in Azure SQL, PostgreSQL, MySQL, or Cosmos DB as typed REST and GraphQL endpoints, with role-based access control and caching handled for you. The MCP support is what makes it interesting for AI work, since the same database becomes a controlled set of tools an agent can call directly, instead of relying on flaky natural-language-to-SQL prompts. It runs as a container on any cloud or on-premises, so it works as a lightweight CRUD service without locking you into a host. For small teams that want a config-driven data layer rather than a big custom service, it's quick to stand up and evaluate.