Google Docs connector

The google-docs MCP adapter wraps the Google Docs API and Drive search. Use it in workflow mcp_call steps to read document text, append LLM-generated sections, find docs by name, and produce filled copies from templates with placeholder replacement. Documents are addressed by Drive file ID.

Prerequisites

A Google account with Google Docs and Drive access
A Google Workspace connection in AgentRuntime (OAuth via the Google account card on Connections)
project_contributor access in your workspace

Connect in AgentRuntime

Connect Google

On Connections, use the Google account card and complete OAuth. See Google Workspace setup and Connections.

Enable Docs

On your Google connection, enable the Docs service so the account exposes Docs and Drive API scopes.

Add google-docs MCP instance

Go to MCP, click Add instance for google-docs. Open Instance config, wire your Google connection, and set the profile active.

Test in a workflow

Smoke-test with mcp_call in Workflow Studio. Test with find_document or read_document on a known doc before production workflows.

What you can build

Contract from template — CRM webhook with deal fields → create_document_based_on_template replacing [ClientName] placeholders → google-drive add_permission for the signer.
Meeting notes append — google-calendar event ends → LLM summary from transcript → find_document by meeting title → append_text with dated section.
Weekly report draft — Postgres metrics query → LLM narrative → create_document with title and body → gmail link in status email.
Onboarding packet — New hire row in google-sheets → create_document_based_on_template with role-specific values → move to HR folder via google-drive.

Tools

Tool	Description
`read_document`	Read a Google Doc by document ID and return both the full resource and flattened plain text.
`append_text`	Append text to the end of an existing Google Doc.
`find_document`	Find a Google Doc by name in Drive, with optional create-if-missing behavior.
`create_document_based_on_template`	Create a Google Doc from a template by replacing placeholders with values and images.
`create_document`	Create a Google Doc with a title and optional initial body content.

Example

Fill a proposal template after an LLM extracts fields from a brief:

{
  "id": "fill-proposal",
  "type": "mcp_call",
  "name": "Generate proposal doc",
  "tool_name": "create_document_based_on_template",
  "tool_args": {
    "template": "{{input.proposal_template_id}}",
    "placeholder_format": "square_brackets",
    "values": {
      "ClientName": "{{input.client_name}}",
      "ProjectScope": "{{steps.llm-scope.result.summary}}",
      "EstimatedCost": "{{input.budget}}"
    }
  },
  "depends_on": ["llm-scope"],
  "timeout_s": 60
}

Troubleshooting

Issue	Fix
Reconnect prompt on Google connection	Re-authorize OAuth; Docs and Drive scopes may have been revoked
Placeholders not replaced	Match `placeholder_format` to your template (`square_brackets`, `curly_braces`, etc.)
`find_document` returns multiple matches	Scope with `folder_id` or use an exact template ID instead of name search
`append_text` lands in wrong doc	Pass the `document_id` from `find_document` or `create_document`, not the URL slug alone

Get started

Platform

Workflows

Guides

Integration setup

Connectors

AI & chat

API

Prerequisites

Connect in AgentRuntime

What you can build

Tools

Example

Troubleshooting

​Prerequisites

​Connect in AgentRuntime

​What you can build

​Tools

​Example

​Troubleshooting

​Related

Prerequisites

Connect in AgentRuntime

What you can build

Tools

Example

Troubleshooting

Related