Skip to main content

Connect Storydoc via MCP to ChatGPT, Gemini, and other AI tools

Step-by-step setup for connecting Storydoc to ChatGPT, Gemini, GitHub Copilot, Microsoft 365 Copilot, and any other AI tool that supports the Model Context Protocol (MCP).

Written by Jack

The Storydoc MCP server lets you bring Storydoc into any AI assistant or platform that supports the Model Context Protocol (MCP) — the open standard most major AI tools have adopted for connecting external tools.

This guide covers the tools that aren't yet on our main connector list but can be set up manually today, including ChatGPT (Developer Mode), Gemini, GitHub Copilot, Microsoft 365 Copilot for organizations, and any other MCP-aware tool.

Audience: This guide is for developers, IT administrators, and power users. If you're connecting Storydoc through Claude (Desktop, Code, or the web), see our Storydoc + Claude guide.

The connection details (same for every tool)

You'll need these in every setup below:

  • MCP Server URL: https://mcp.storydoc.com/mcp

  • Authentication: OAuth (you'll paste your Storydoc API token on a sign-in page)

  • API token: generate one at Profile → Integrations → Automations in your Storydoc account

Keep that token handy — you'll paste it once per tool.

Connect to ChatGPT (Developer Mode)

Storydoc is going through OpenAI's official review for the ChatGPT Apps Directory. Until that's live, you can connect it today through ChatGPT's Developer Mode — available on Plus, Pro, and Team plans.

Steps:

  1. In ChatGPT, go to Settings → Connectors.

  2. Toggle on Developer Mode.

  3. Click + Add custom connector.

  4. Enter:

    • Name: Storydoc

    • MCP Server URL: https://mcp.storydoc.com/mcp

    • Authentication: OAuth

  5. Click Create, then complete the Storydoc sign-in by pasting your API token.

What happens next: Storydoc appears alongside your other connectors. Try a prompt like "Create a Storydoc proposal for Acme Corp, recipient Sarah Chen." ChatGPT will list your templates, ask for the missing variables, and return a shareable link.

Once the official Storydoc app is approved in the ChatGPT Apps Directory, you'll be able to install it in one click without Developer Mode. We'll update this article when that happens.

Connect to Gemini

Gemini's web and mobile apps don't currently support custom connectors. The Storydoc MCP works in Gemini CLI (and the new Antigravity CLI, which is replacing Gemini CLI for free-tier users) — a command-line tool aimed at developers and automation.

Best for: scripting Storydoc creation into automated workflows, generating decks as part of CI/CD or data pipelines, and power users who work in a terminal.

Steps (Gemini CLI):

  1. Install Gemini CLI by following Google's setup guide.

  2. Open your Gemini settings file:

    • macOS / Linux: ~/.gemini/settings.json

    • Windows: %USERPROFILE%\.gemini\settings.json

  3. Add Storydoc to the mcpServers section:

    {   "mcpServers": {     "storydoc": {       "url": "https://mcp.storydoc.com/mcp"     }   } }

  4. Save and start Gemini CLI by running gemini in your terminal.

  5. Trigger the Storydoc sign-in:

    /mcp auth storydoc

    A browser opens. Paste your Storydoc API token to authorize, then return to the terminal.

  6. Confirm it's connected:

    /mcp list

    You should see storydoc - Connected with six tools listed.

What you can do with it: anything you'd do with Storydoc in chat — "list my templates," "create a proposal from this template for…," "pull analytics for the Nike deck" — but executed from a terminal or scripted into your own workflows. This makes it useful for automation: triggering Storydoc creation from data in a spreadsheet, scheduled reporting, or as part of a larger agentic pipeline.

Gemini limitations to know

  • Web and mobile Gemini do not support custom MCP connectors today. This is a Google product limitation, not a Storydoc one. When Google opens a consumer connector directory, Storydoc will be available there.

  • Gemini CLI is being replaced by Antigravity CLI for free and individual users. The setup is nearly identical — same settings.json pattern, same Storydoc URL — but you'll need to install Antigravity CLI separately. Google's migration page has details.

  • The Storydoc sign-in must be triggered manually via /mcp auth storydoc. Other clients prompt automatically; Gemini doesn't yet.

Connect to GitHub Copilot (VS Code, JetBrains)

GitHub Copilot supports MCP servers through your IDE's settings file. Any developer with Copilot access can add Storydoc personally — no admin required.

VS Code:

  1. Open VS Code settings (JSON view).

  2. Add an mcp section:

    {   "mcp": {     "servers": {       "storydoc": {         "type": "http",         "url": "https://mcp.storydoc.com/mcp"       }     }   } }

  3. Save. Copilot Chat will prompt you to authorize Storydoc the first time you invoke a Storydoc tool.

JetBrains IDEs: follow the same pattern in your Copilot configuration file — the JSON shape is identical.

Best for: developers building Storydoc workflows directly inside their IDE, or asking Copilot Chat to create decks as part of a development task.

Connect to Microsoft 365 Copilot (for IT administrators)

Microsoft 365 Copilot — the assistant inside Word, Outlook, Teams, and PowerPoint — supports custom MCP servers through Copilot Studio. This is an admin-managed integration: end users can't self-install, but IT can deploy Storydoc once and make it available org-wide.

Steps for IT administrators:

  1. Open Microsoft Copilot Studio with an account that has Copilot Studio admin permissions.

  2. Open or create the Copilot agent where you want Storydoc available.

  3. Go to Tools → Add a tool → Model Context Protocol (MCP).

  4. Enter:

    • Server name: Storydoc

    • Server URL: https://mcp.storydoc.com/mcp

    • Authentication: OAuth 2.0

  5. Complete the OAuth setup. Copilot Studio guides you through registering the connection and configuring the consent flow for your users.

  6. Publish the agent. The Storydoc tools become available to users assigned to that agent inside their Microsoft 365 Copilot experience.

What your users get: when they ask Copilot in Word, Outlook, or Teams to "create a Storydoc proposal for [client]" or "check how the [deck] is performing," the agent will use the Storydoc tools you connected.

Notes for IT teams:

  • Each user authenticates against Storydoc individually using their own API token the first time they use a Storydoc tool — their decks stay scoped to their own Storydoc account.

  • Storydoc API tokens are generated from each user's Storydoc account under Profile → Integrations → Automations. Plan a short rollout note for your users so they know where to find theirs.

  • For broader Microsoft 365 distribution (a published app in AppSource), Storydoc is exploring an official Microsoft 365 listing. This admin-managed path is available today and doesn't require waiting for that listing.

Connect to any other MCP-aware tool

The MCP standard is being adopted across the AI industry. If your team uses an internal AI tool, a vendor's assistant, or a less common public client, there's a good chance it already supports MCP — or will soon.

How to know if a tool supports MCP

Look for any of the following in the tool's settings, documentation, or admin console:

  • A section called Connectors, Custom Tools, MCP, or Model Context Protocol

  • An option to add a Custom MCP server by URL

  • A configuration file referencing mcpServers (a JSON object mapping a name to a server config)

  • Support for OAuth 2.0 authentication on external tools

If you find any of these, the tool can almost certainly connect to Storydoc.

What to provide

In every MCP-aware tool, you'll need to give it:

  • Server URL: https://mcp.storydoc.com/mcp

  • Transport: Streamable HTTP (this is what hosted MCP servers use; some older tools may also support SSE)

  • Authentication: OAuth 2.0 — the tool will redirect users to a Storydoc page where they paste their API token

That's the complete configuration. There's no installation, no package to download, and no per-tool customization on Storydoc's side.

If your tool uses a config file

Most MCP-aware tools use a JSON config of this shape:

{   "mcpServers": {     "storydoc": {       "url": "https://mcp.storydoc.com/mcp"     }   } }

The exact location varies by tool (~/.gemini/settings.json, VS Code settings, an mcp.json file, etc.), but the contents are the same.

If your tool uses an admin console

Look for Add custom connector or Add MCP server in the admin or workspace settings. Provide the URL above and select OAuth as the auth method.

If you're not sure

If you can't tell whether your tool supports MCP, two quick checks:

  • Search the tool's documentation for "MCP" or "Model Context Protocol."

  • Ask the tool's support team whether they support custom MCP servers.

If the answer is no today, watch for updates — MCP support is being added rapidly across the industry, and Storydoc works on the open standard, so any tool that adopts MCP will work with it automatically.

Need help?

Reach out at [email protected] — happy to help with any specific tool not covered above.

Did this answer your question?