Why pgEdge thinks MCP (not an API) is the right way for AI agents to talk to databases

pgEdge has launched a production-ready MCP Server for Postgres, enabling AI agents and LLMs to connect to any standard Postgres database (v14+) without relying on traditional APIs. Key features include built-in security (HTTPS/TLS, token-based auth, read-only default), full schema introspection covering primary keys, foreign keys, indexes, and constraints, and token usage reductions of 30–50% via TSV output, pagination, and context window compaction. The server works with Claude Code, Cursor, Windsurf, VS Code Copilot, and models from OpenAI, Anthropic, and locally hosted options. It supports on-premises, self-managed cloud, air-gapped environments, and pgEdge Cloud. Custom tools can be written in SQL, Python, Perl, or JavaScript, and a DBA toolkit provides database health analysis and index recommendations. The project is fully open source under the Postgres license.