Database Access Patterns

Database MCP servers give Gemini CLI direct access to your data layer. The official @modelcontextprotocol/server-postgres handles PostgreSQL. Community servers cover MySQL, SQLite, MongoDB, and Redis. Once configured, Gemini can query your database, inspect schemas, generate migration files, and debug data issues — all without leaving the terminal session.

The practical workflow changes when your AI agent can see your data. Instead of describing a bug verbally — "the user count seems wrong on the dashboard" — you can say "query the users table and compare the count to what the /api/users endpoint returns." Gemini runs the database query via the MCP server, fetches the API response via web_fetch or run_shell_command, and identifies the discrepancy. Data-driven debugging replaces guesswork.

{
  "mcpServers": {
    "postgres": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-postgres"],
      "env": {
        "DATABASE_URL": "${DATABASE_URL}"
      }
    },
    "sqlite": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-sqlite", "--db", "./data/app.sqlite"]
    }
  }
}