# PlanetScale MCP Server

> Lets AI clients query and manage PlanetScale databases

- Page on StartupList: https://startuplist.tech/launches/planetscale-mcp-server
- Website: https://planetscale.com/docs/mcp-server
- Category: MCP Servers
- Price: $5/mo
- Company: PlanetScale MCP Server (https://startuplist.tech/companies/planetscale-mcp-server)
- Used for: Analyse data, Automate workflows

## At a glance

- The problem it solves: Developers have no direct way to let AI assistants inspect and query their PlanetScale databases from the tools they already use
- Who it's for: Businesses (B2B)
- How it finds customers: Blog and content, Marketplaces and integrations

## About

The PlanetScale MCP server is a hosted Model Context Protocol server that connects AI tools such as Claude, Cursor, VS Code, Codex and Notion to your PlanetScale organizations, databases, branches, schema, Insights data and organization payment method.

It is aimed at developers and teams who already run PlanetScale databases and want an AI assistant to work with them. Depending on the permissions you grant, a client can read schema and Insights data, run read queries, and run write queries. An insights-only server is also available for users who want Insights and Schema Recommendations without any query execution.

You connect a client by adding the server URL, https://mcp.pscale.dev/mcp/planetscale, and signing in with PlanetScale through OAuth. Each client registers as its own OAuth application. For CI jobs and headless agents, PlanetScale offers MCP service tokens instead. The page includes setup steps for Cursor, VS Code, Claude.ai and Claude for desktop, Claude Code, OpenCode, Codex CLI, Gemini CLI, Amp CLI and Notion.

Access is set in layers. You choose no, read-only or full access per organization or per database, and organization administrators can limit query access for all members. Each query uses short-lived credentials that are deleted after it runs. The page advises caution with write access to production databases and recommends reviewing queries before they run.

The server is free to connect as far as the page states, but the databases it works with are billed separately. PlanetScale's Base plan is self-serve and bills for cluster sizes, storage and extra production branches, with some Postgres clusters starting at $5 a month. Enterprise plans are available for larger needs.

## Links

- [PlanetScale MCP Server on StartupList](https://startuplist.tech/launches/planetscale-mcp-server)
- [Website](https://planetscale.com/docs/mcp-server)