Skip to content
PlanetScale

PlanetScaleMCP Server

by PlanetScale
Released
v1.0.0
Apache-2.0
Repository Docs
Databases & Data

Summary

PlanetScale's official hosted MCP server — browse organizations, databases and branches, read schema, run guarded SQL, and pull Insights and schema recommendations.

Features

  • Discover organizations, databases and branches
  • Read branch schema and list schema recommendations
  • Run read and write SQL with per-query ephemeral credentials
  • Blocks UPDATE/DELETE without WHERE and TRUNCATE; DDL needs confirmation
  • Query Insights performance data for slow and expensive queries
  • Insights-only endpoint that removes query execution entirely
  • Billing tools: invoices, line items and cluster size SKUs
  • OAuth for interactive clients, service tokens for CI and headless agents

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it.
  3. Restart the agent — the "PlanetScale" tools will be available.
{
  "mcpServers": {
    "planetscale": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.pscale.dev/mcp/planetscale"
      ]
    }
  }
}

Description

PlanetScale MCP is the company's own hosted Model Context Protocol server, at https://mcp.pscale.dev/mcp/planetscale. It gives an agent an OAuth-scoped view of your PlanetScale account: which organizations and databases exist, what branches they have, what the schema on a branch actually looks like, and — the part that matters when you are debugging — what Insights says about how those queries have been performing.

Tools. Sixteen in the hosted server, generated in part from the PlanetScale OpenAPI spec. They cover discovery (planetscale_list_organizations, planetscale_list_databases, planetscale_list_branches, planetscale_get_database, planetscale_get_branch), schema (planetscale_get_branch_schema, planetscale_list_schema_recommendations), query execution (planetscale_execute_read_query, planetscale_execute_write_query), performance (planetscale_get_insights), billing (planetscale_list_invoices, planetscale_get_invoice_line_items, planetscale_list_cluster_size_skus) and planetscale_search_documentation.

The safeguards are the interesting design decision. Every query runs with short-lived ephemeral credentials that are created on demand and deleted immediately after execution, so no long-lived database password ends up in a config file. UPDATE and DELETE without a WHERE clause are blocked outright, TRUNCATE is blocked, and DDL requires human confirmation. If you want none of that risk, there is a second endpoint — https://mcp.pscale.dev/mcp/planetscale-insights-only — that drops both query-execution tools and leaves you with Insights and schema recommendations alone. That read-only variant is the sane default for handing an agent access to a production database.

Auth and clients. OAuth for interactive clients; MCP service tokens for CI, headless agents and anything that can set a custom HTTP header. One-click installs exist for Cursor and VS Code, with documented setups for Claude Code, Claude Desktop and claude.ai via the Connectors directory, OpenCode, Codex CLI, Amp CLI, Gemini CLI and Notion. The open-source portion of the server is Apache-2.0 at planetscale/mcp-server.

Related MCP Servers

MCP: Tableau MCP

by Tableau

New

Tableau's own MCP server — query published data sources, explore workbooks and views, and render live vizzes inside an AI client.

Databases & Data
Browse all MCP servers →