A1-x-Tech/mcp-google-slides

English | Русский

npm Glama CI License: MIT

A1 Google Slides MCP lets an AI app build and edit Google Slides presentations in plain language. Draft a deck, add and rearrange slides, write and style text, place shapes, images and tables, keep speaker notes, review comments and export the result as PDF or PPTX.

It uses the Google Slides API with your Google account. Everything inside a deck is addressed by explicit object ids, and it makes the limits of the Slides API explicit instead of implying that every presentation task is possible.

• 32 tools. Inspect decks and pages, edit slides, text, shapes, images and tables, manage speaker notes and comments, render thumbnails and export files.
• Connects from the conversation. Say "connect Google Slides": the server walks you through the OAuth client, catches Google's redirect on 127.0.0.1 with PKCE and keeps the tokens itself — no config files, no restart.
• Edits are atomic. Changes ride the API's batchUpdate: one invalid request voids the whole batch, so a deck is never left half-edited.
• Your Drive stays out of reach. Comments, export and local-image upload use the Drive API strictly for the one presentation file — there are no tools to list, share, rename or delete Drive files.
• Minimal Google scopes. It needs presentations; Drive scopes are required only for local images, comments and export.

Start with a read-only question:

Show me the slides in the quarterly report deck and what's on each of them.

Connect the server · Explore use cases · Open technical documentation

See it work in a minute

You: List the slides of the Q3 review deck and what's on each one.

Assistant: Shows a compact inventory: every slide's position, layout and elements with their visible text. Nothing changes.

You: Prepare a new slide after the agenda titled “Roadmap”, with the key dates in the body.

Assistant: Shows the target deck, the layout and the proposed text, then asks for confirmation before adding it.

You: Confirm.

Assistant: Adds the slide and fills its placeholders. It does not touch the other slides.

Contents

• Quick start
• What you can ask it to do
• How a presentation changes
• What can change
• Getting access
• Configuration
• Data, limits and background work
• Technical documentation
• Support

Quick start

You need Node.js 20+ and a Google account. Credentials are not required at install time — the server connects from the conversation.

1. Add the server to your AI app.
2. Say "connect Google Slides": the assistant walks you through creating the OAuth client and approving access without editing config files.
3. Ask the read-only question above.

codex mcp add google-slides \
  -- npx -y @a1-x-tech/mcp-google-slides@latest

codex mcp list

claude mcp add \
  --transport stdio --scope user google-slides \
  -- npx -y @a1-x-tech/mcp-google-slides@latest

claude mcp list

{
  "mcpServers": {
    "google-slides": {
      "command": "npx",
      "args": ["-y", "@a1-x-tech/mcp-google-slides@latest"]
    }
  }
}

{
  "mcpServers": {
    "google-slides": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@a1-x-tech/mcp-google-slides@latest"]
    }
  }
}

{
  "servers": {
    "google-slides": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@a1-x-tech/mcp-google-slides@latest"]
    }
  }
}

What you can ask it to do

Inspect a deck

• List the slides of this presentation and what's on each one.
• Read the speaker notes for slide 5.
• Render a thumbnail of the title slide.

Build and rearrange slides

• Create a presentation for the Q3 review.
• Add a “Roadmap” slide after the agenda, duplicate the template slide, move the summary to the end.
• Insert an image from a URL or a local file, add a 4×3 table, create a text box or a rounded rectangle with its caption.

Write and polish text

• Put this text into the title placeholder and make the deadline bold and red.
• Replace “Q2” with “Q3” across the whole deck.
• Rewrite the speaker notes of slide 3 to be shorter.

Review and hand off

• List the open comment threads and reply to the first one, resolving it.
• Delete the draft slide we no longer need.
• Export the deck as PDF or PPTX for someone without Google access.

How a presentation changes

1. create_presentation creates a deck with one title slide; only the title is honored at creation.
2. Everything inside a deck — slides, shapes, images, tables — is addressed by object ids. list_slides is the compact inventory to start from before any edit.
3. Edits go through the API's atomic batchUpdate: one invalid request voids the whole batch, so nothing is half-applied.
4. Tool inputs are in points; the API stores EMU (1 pt = 12700 EMU). The server converts both ways.

The Slides API cannot delete, rename or share the presentation file — those are Drive file operations outside this server's scope. It cannot import a theme (only instantiate layouts the deck already has) and cannot anchor a new comment to a specific slide. Export is capped at 10 MB, and thumbnail URLs expire after about 30 minutes.

What can change

The AI client controls confirmation prompts. The server marks reads, writes and destructive tools so the client can distinguish an inspection from a live change.

Getting access

Google Slides requires OAuth 2.0; an API key is not enough. There are two ways in, and the first one needs no configuration files.

Connect from the chat (recommended)

Say "connect Google Slides" and the assistant runs the flow with you:

1. setup_instructions prints the checklist: create or select a Google Cloud project, enable Google Slides API, configure the consent screen and create a Desktop app OAuth client.
2. Download that client's JSON ("Download JSON") and give the assistant its path — set_client stores it owner-only. The secret never goes through the conversation.
3. start_login returns a Google consent link. Open it on this machine and approve; the code comes back to a one-shot listener on 127.0.0.1 (PKCE), never through the chat.
4. finish_login exchanges the code and saves the tokens to ~/.config/mcp-google-slides/credentials.json (mode 0600).

The tokens are re-read on every call, so the connection works immediately — no restart of the AI app. auth_status shows what is connected, logout revokes and deletes it.

Environment variables (CI, unattended installs)

1. Create or select a Google Cloud project and enable Google Slides API. Enable Google Drive API too if you want comments, export or local-image upload.

2. Configure the OAuth consent screen and create a Desktop app OAuth client.

3. Authorize the Google account that owns or can edit the presentations. The OAuth 2.0 Playground can obtain the refresh token when Use your own OAuth credentials is enabled.

4. Request the scopes:

   https://www.googleapis.com/auth/presentations
   https://www.googleapis.com/auth/drive.file
   

   presentations covers all Slides reads and writes. drive.file covers local-image upload plus comments and export on decks created through this app; request the broader https://www.googleapis.com/auth/drive instead when you need comments or export on arbitrary presentations.

Testing-mode OAuth refresh tokens can expire after seven days. Publish the OAuth app, or use an Internal app in a Workspace domain, when you need long-lived access. Treat the client secret and refresh token as passwords.

Configuration

Every variable is optional — with none of them the server connects from the chat.

* Provide either the OAuth trio or an access token.

Data, limits and background work

• Requests go to Google Slides. The local server refreshes Google OAuth tokens and calls the Slides API; comments, export and local-image upload ride the Drive API, scoped to the one presentation file. Its anonymous telemetry contains an installation ID, package version, AI client and platform versions, and tool names — never OAuth tokens, presentation content, tool arguments or prompts. For each tool call it also sends the outcome and, on failure, an excerpt of the error text with e-mails, addresses, tokens, numbers of 4+ digits and quoted text replaced by #; a value from your request can appear in it if the error text contains it. Set ASKADS_TELEMETRY=0 to opt out.
• Google applies rate limits. On 429, the server uses backoff; reads also retry after network and 5xx errors, while writes are not replayed after an uncertain failure. Thumbnails count against a more expensive read quota.
• There is no background polling. The server runs only when called. Nothing watches a deck between calls; if your AI app supports scheduled tasks, it can check a presentation periodically.

Technical documentation

• MCP capability catalog — task-oriented pages for every tool.
• All tools and inputs
• Development documentation
• Publishing documentation
• Google Slides API reference

Support

Found a bug or need a scenario? Create an issue or write in Telegram.

You made it to the end!