> ## Documentation Index
> Fetch the complete documentation index at: https://docs.plyne.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect ChatGPT to Plyne

> Add a custom MCP server in ChatGPT and authorize access to your saved Plyne results.

Connect Plyne through ChatGPT's custom MCP server setup. Your account and workspace must permit custom MCP connections; a workspace administrator may need to create or approve the connection.

Use this public server address. You will sign in to Plyne when authorizing access to your workspace.

```text Server URL theme={null}
https://api.plyne.ai/mcp
```

**1. Add a custom MCP server**

In ChatGPT, open **Plugins**, select the plus button, and choose **Add custom MCP server**. Name it **Plyne**, paste the URL above, and choose **OAuth** authentication. Use the client-registration option supported by the detected authorization server. No API key is needed.

**2. Approve the workspace**

Follow the browser authorization flow to Plyne. Check the application details, choose the intended workspace, and approve the requested read access.

**3. Test the connection**

Select the Plyne connection in a conversation and ask: **“List my Signal Agents.”** A successful setup should return your workspace's results, including an empty list when no agents exist.

Workspace policies, publication controls, and client labels can vary. Follow OpenAI's [current connection guide](https://help.openai.com/en/articles/12584461-connecting-custom-mcp-servers-in-chatgpt) if your interface differs.

Plyne's MCP tools read saved records and evidence. They do not change records or send messages. Manage or remove the Plyne connection from ChatGPT’s connector settings.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.