Skip to main content
Control Plane is a listed plugin in the ChatGPT Plugin directory, so there is no MCP URL to enter by hand: open the listing, click Install plugin, and authorize your orgs.
The same listing reaches ChatGPT on the web, desktop, and mobile, Codex in the ChatGPT desktop app, the Codex IDE extension (it shows up under Codex settingsPlugins, installed and enabled), and Codex CLI through its /plugins browser. See OpenAI Codex.

Prerequisites

ChatGPT account

Sign in at chatgpt.com

Control Plane account

Sign up at console.cpln.io if you don’t have an account yet

Install the Control Plane plugin

1

Open the Control Plane listing

Open the Control Plane plugin listing.
If that link doesn’t open the listing, click Plugins in the left sidebar and search for Control Plane.
2

Install plugin

Click Install plugin. Installation starts the authorization flow.
3

Authorize your orgs

When ChatGPT prompts you to Connect:
  1. Sign in to Control Plane (Google, GitHub, Microsoft, or SAML).
  2. Select the orgs you want ChatGPT to access.
  3. Click Authorize.
4

Use it

Start a new chat. Describe what you want and ChatGPT picks the plugin up on its own, or type @ to invoke Control Plane explicitly. Ask it to “list GVCs in org <your-org>” (using one of the orgs you authorized) to confirm.
The listing is fixed on the default core toolset profile and takes no query parameter of its own. For mk8s or full, register the MCP server directly in Codex or another MCP client. See toolset profiles.
For the latest directory behavior, see OpenAI’s Plugins documentation.

Example conversation

1

Set Context

2

Deploy a Workload

3

Get the Endpoint

Troubleshooting

Open the Control Plane plugin listing directly. If it still doesn’t install, admins on ChatGPT Business, Enterprise, and Edu workspaces choose which plugins their members can access under AdminPlugins, so ask your administrator to allow Control Plane.
Bundled tools become available in a new chat. Start a fresh conversation, then type @ and pick Control Plane to invoke it explicitly.
Click Plugins in the left sidebar, open Control Plane, disconnect it, then connect again. Sign in and adjust your org selections on the consent screen.
A profile is fixed for the life of a connection, so reconnect on a wider one: point an MCP client at https://mcp.cpln.io/mcp?toolsets=full.

Next steps

OpenAI Codex

Set up Codex CLI and the Codex IDE extension

Usage Examples

Practical prompts for common Control Plane workflows

Tools Reference

Every tool the MCP server exposes, grouped by resource

Template Catalog

30+ production-ready templates your AI can deploy on request