The Gemini CLI confirmed that the MCP server built with Google Apps Script (GAS), a low-code platform, offers immense possibilities. If you've created snippets for GAS, these could be revitalized and/or leveraged in new ways by using them as the MCP server. The Gemini CLI and other MCP clients will be useful in achieving this.

The Gemini CLI provides a powerful command-line interface for interacting with Google's Gemini models. By leveraging the Model Context Protocol (MCP), the CLI can be extended with custom tools. This report explores the integration of the Gemini CLI with an MCP server built using Google Apps Script Web Apps. We demonstrate how this combination simplifies authorization for Google Workspace APIs (Gmail, Drive, Calendar, etc.), allowing Gemini to execute complex, multi-step tasks directly within the Google ecosystem. We provide setup instructions and several practical examples showcasing how this integration unlocks significant potential for automation and productivity enhancement.

Recently, I published a report titled "Gemini CLI with MCP Server Built by Web Apps of Google Apps Script" (Ref). This initial report highlighted how a Model Context Protocol (MCP) server, developed using Google Apps Script Web Apps, can be integrated with the Gemini CLI.

One of the key advantages of using Google Apps Script is its ability to provide seamless authorization for Google APIs, particularly those within Google Workspace. This significantly reduces the development overhead associated with building applications that interact with Google Workspace services.

Building upon that foundation, this report aims to explore the expanded possibilities when combining the Gemini CLI with an MCP server powered by Google Apps Script Web Apps. We will delve into how this powerful combination facilitates the integration of various tools and services, unlocking an infinite range of potential applications for enhanced productivity and automation within the Google ecosystem.

In the current stage (August 22, 2025), the following 125 tools are provided by ToolsForMCPServer for the MCP server.

• add_label_to_Gmail: Add labels to threads of Gmail. Don't use the invalid thread IDs.
• analytics_admin_accountSummaries_list: Use to retrieve a list of all Google Analytics accounts accessible by the current user. Each entry provides key details for the account and a summary of its properties, making it useful for discovering available data streams and managing permissions.
• analytics_admin_properties_get: Use to get detailed information about a single Google Analytics property, providing essential details for management and analysis.
• analytics_data_properties_runRealtimeReport: Use to generate a customized report of real-time event data from a Google Analytics property, showing events and user activity that occurred within the last 30 minutes. Useful for monitoring live traffic and immediate user behavior.
• analytics_data_properties_runReport: Use to fetch a custom report from a Google Analytics property.
• auto_new_draft_creation_Gmail: Create automatically drafted emails in Gmail. This function returns the value, including the message ID and the draft ID. When creating the draft email, confirm the owner's name and insert the sender's name into the footer. Don't use '[Your Name]'. If you have no information about the sender's email, don't include the footer of sender's name in the email.
• auto_reply_draft_creation_Gmail: Create automatically drafted reply emails in Gmail. Don't use the invalid message IDs. This function returns the value including the message ID, the draft ID, the URL of the mail.
• change_permission_of_file_on_google_drive: Use to change the permission of a file or folder on Google Drive for a specific user by providing the item ID, user email, and desired role. As a sample situation, when URLs of the files are included in an email, it is required to add the permission to the recipient user to allow the user to read or write the file.
• classroom_courses_aliases_create: Use to create an alias for a course.
• classroom_courses_aliases_delete: Use to delete an alias of a course.
• classroom_courses_aliases_list: Use to get a list of aliases.
• classroom_courses_announcements_create: Use to creates an announcement using the "courses.announcements.create" method of Google Classroom API.
• classroom_courses_announcements_delete: Use to delete an announcement.
• classroom_courses_announcements_get: Use to return an announcement.
• classroom_courses_announcements_list: Use to return a list of announcements that the requester is permitted to view.
• classroom_courses_announcements_modifyAssignees: Use to modify assignee mode and options of an announcement using the "courses.announcements.modifyAssignees" method of Google Classroom API.
• classroom_courses_announcements_patch: Use to update one or more fields of an announcement using the "courses.announcements.patch" method of Google Classroom API.
• classroom_courses_courseWorkMaterials_create: Use to creates a course work material using the "courses.courseWorkMaterials.create" method of Google Classroom API.
• classroom_courses_courseWorkMaterials_delete: Use to delete a course work material.
• classroom_courses_courseWorkMaterials_get: Use to return a course work material.
• classroom_courses_courseWorkMaterials_list: Returns a list of course work material that the requester is permitted to view.
• classroom_courses_courseWorkMaterials_patch: Use to update one or more fields of a course work material using the "courses.courseWorkMaterials.patch" method of Google Classroom API.
• classroom_courses_courseWork_create: Use to create course work using the "courses.courseWork.create" method of Google Classroom API.
• classroom_courses_courseWork_delete: Use to delete a course work.
• classroom_courses_courseWork_get: Use to return metadata of course work.
• classroom_courses_courseWork_list: Use to return a list of course work that the requester is permitted to view.
• classroom_courses_courseWork_modifyAssignees: Use to modify assignee mode and options of a coursework.
• classroom_courses_courseWork_patch: Use to update one or more fields of a course work using the "courses.courseWork.patch" method of Google Classroom API.
• classroom_courses_courseWork_rubrics_create: Use to creates a rubric using the "courses.courseWork.rubrics.create" method of Google Classroom API.
• classroom_courses_courseWork_rubrics_delete: Use to delete a rubric using the "courses.courseWork.rubrics.delete" method of Google Classroom API.
• classroom_courses_courseWork_rubrics_get: Use to return a rubric using the "courses.courseWork.rubrics.get" method of Google Classroom API.
• classroom_courses_courseWork_rubrics_list: Returns a list of rubrics that the requester is permitted to view.
• classroom_courses_courseWork_rubrics_patch: Use to update a rubric using the "courses.courseWork.rubrics.patch" method of Google Classroom API.
• classroom_courses_courseWork_studentSubmissions_get: Use to return the metadata of a student submission using the "courses.courseWork.studentSubmissions.get" method of Google Classroom API.
• classroom_courses_courseWork_studentSubmissions_list: Returns a list of student submissions that the requester is permitted to view.
• classroom_courses_courseWork_studentSubmissions_patch: Use to update one or more fields of a student submission using the "courses.courseWork.studentSubmissions.patch" method of Google Classroom API.
• classroom_courses_courseWork_studentSubmissions_reclaim: Use to reclaim a student submission on behalf of the student that owns it using the "courses.courseWork.studentSubmissions.reclaim" method of Google Classroom API.
• classroom_courses_courseWork_studentSubmissions_return: Use to returns a student submission using the "courses.courseWork.studentSubmissions.return" method of Google Classroom API.
• classroom_courses_courseWork_studentSubmissions_turnIn: Use to turn in a student submission using the "courses.courseWork.studentSubmissions.turnIn" method of Google Classroom API.
• classroom_courses_create: Use to create a course using the "courses.create" method of Google Classroom API. The user specified in ownerId is the owner of the created course and added as a teacher. A non-admin requesting user can only create a course with themselves as the owner. Domain admins can create courses owned by any user within their domain.
• classroom_courses_get: Use to return metadata of a course.
• classroom_courses_getGradingPeriodSettings: Use to return the grading period settings in a course.
• classroom_courses_list: Use to retrieve courses of Google Classroom using a method "courses.list" of Google Classroom API. Unless otherwise specified, run this tool without parameters of "studentId", "teacherId", and "courseStates".
• classroom_courses_patch: Use to update one or more fields in a course using the "courses.patch" method of Google Classroom API.
• classroom_courses_remove: Use to deletes a course using the "courses.delete" method of Google Classroom API.
• classroom_courses_students_create: Use to add a user as a student of a course. using the "courses.students.create" method of Google Classroom API.
• classroom_courses_students_delete: Use to delete a student of a course.
• classroom_courses_students_get: Use to return a student of a course.
• classroom_courses_students_list: Use to return a list of students of this course that the requester is permitted to view.
• classroom_courses_teachers_create: Use to add a user as a teacher of a course using the "courses.teachers.create" method of Google Classroom API.
• classroom_courses_teachers_delete: Use to delete a teacher of a course.
• classroom_courses_teachers_get: Use to return a teacher of a course.
• classroom_courses_teachers_list: Use to return a list of teachers of this course that the requester is permitted to view.
• classroom_courses_topics_create: Use to creates a topic using the "courses.topics.create" method of Google Classroom API.
• classroom_courses_topics_delete: Use to delete a topic.
• classroom_courses_topics_get: Use to return a topic.
• classroom_courses_topics_list: Use to return a list of topics that the requester is permitted to view.
• classroom_courses_topics_patch: Use to update one or more fields of a topic using the "courses.topics.patch" method of Google Classroom API.
• classroom_courses_update: Use to update a course using the "courses.update" method of Google Classroom API.
• classroom_courses_updateGradingPeriodSettings: Use to update grading period settings of a course using the "courses.updateGradingPeriodSettings" method of Google Classroom API.
• classroom_invitations_accept: Use to accept an invitation using the "invitations.accept" method of Google Classroom API. Accepts an invitation, removing it and adding the invited user to the teachers or students (as appropriate) of the specified course. Only the invited user may accept an invitation.
• classroom_invitations_create: Use to create an invitation using the "invitations.create" method of Google Classroom API. Only one invitation for a user and course may exist at a time. Delete and re-create an invitation to make changes.
• classroom_invitations_get: Use to return an invitation using the "invitations.get" method of Google Classroom API.
• classroom_invitations_list: Use to retrieve a list of invitations that the requesting user is permitted to view, restricted to those that match the list request of Google Classroom using a method "invitations.list" of Google Classroom API. At least one of userId or courseId must be supplied. Both fields can be supplied.
• classroom_invitations_remove: Use to deletes an invitation using the "invitations.delete" method of Google Classroom API.
• classroom_registrations_create: Use to create a Registration using the "registrations.create" method of Google Classroom API. Creates a Registration, causing Classroom to start sending notifications from the provided feed to the destination provided in cloudPubSubTopic. Returns the created Registration. Currently, this will be the same as the argument, but with server-assigned fields such as expiryTime and id filled in. Note that any value specified for the expiryTime or id fields will be ignored. While Classroom may validate the cloudPubSubTopic and return errors on a best effort basis, it is the caller's responsibility to ensure that it exists and that Classroom has permission to publish to it.
• classroom_registrations_delete: Use to deletes a Registration using the "registrations.delete" method of Google Classroom API. Deletes a Registration, causing Classroom to stop sending notifications for that Registration.
• classroom_userProfiles_get: Use to return a user profile using the "userProfiles.get" method of Google Classroom API.
• classroom_userProfiles_guardianInvitations_create: Use to creates a guardian invitation using the "courses.topics.create" method of Google Classroom API. Creates a guardian invitation, and sends an email to the guardian asking them to confirm that they are the student's guardian.
• classroom_userProfiles_guardianInvitations_get: Use to return a specific guardian invitation using the "userProfiles.guardianInvitations.get" method of Google Classroom API.
• classroom_userProfiles_guardianInvitations_list: Use to retrieve a list of guardian invitations that the requesting user is permitted to view, filtered by the parameters provided using a method "userProfiles.guardianInvitations.list" of Google Classroom API.
• classroom_userProfiles_guardianInvitations_patch: Use to modify a guardian invitation using the "userProfiles.guardianInvitations.patch" method of Google Classroom API. Currently, the only valid modification is to change the state from PENDING to COMPLETE. This has the effect of withdrawing the invitation.
• classroom_userProfiles_guardians_get: Use to return a specific guardian using the "userProfiles.guardians.get" method of Google Classroom API.
• classroom_userProfiles_guardians_list: Use to retrieve a list of guardians that the requesting user is permitted to view, restricted to those that match the request using a method "userProfiles.guardianInvitations.list" of Google Classroom API. To list guardians for any student that the requesting user may view guardians for, use the literal character - for the student ID.
• classroom_userProfiles_guardians_remove: Use to delete a guardian using the "userProfiles.guardians.delete" method of Google Classroom API.
• comments_drive_api_list: Use to get a list of a file's comments on Google Drive.
• comments_drive_api_remove: Use to delete a comment using the "comments.delete" method of Google Drive API.
• convert_mimetype_of_file_on_google_drive: Use this to convert the mimeType of files on Google Drive.
• create_file_to_google_drive: Use this to create an empty file to Google Drive.
• create_google_docs_from_markdown_on_google_drive: Use to create a Google Document from a markdown format.
• create_schedule_on_Google_Calendar: Use to create a new schedule (event) on Google Calendar.
• delete_schedules_on_Google_Calendar: Use to delete schedules (events) from Google Calendar.
• description_video_on_youtube: Use this to describe and summarize a video on YouTube using the YouTube URL.
• description_web_site: Use this to describe sites using URLs.
• drive_activity_api_query: Use to query past activity in Google Drive. The activities of the files and folders in Google Drive are retrieved.
• generate_description_on_google_drive: Set a description to the file on Google Drive. Use this to generate the description of the file and set it to the file on Google Drive.
• generate_image_on_google_drive: Use this to generate an image from an inputted prompt. The generated image is saved as a file on Google Drive.
• generate_presentation_with_google_slides: Use this to create and generate a presentation using Google Slides.
• generate_quiz_with_google_forms: Use this to generate a quiz with Google Forms. If the number of total questions is not provided, please create 5 questions as the default number of questions.
• generate_roadmap_to_google_sheets: This generates a roadmap in Google Sheets. Use this to generate a roadmap to Google Sheets. Spreadsheet ID and your goal of the roadmap are required to be provided.
• generate_survey_with_google_forms: Use this to generate a survey with Google Forms. If the number of total questions is not provided, please create 5 questions as the default number of questions.
• get_attachment_files_from_Gmail: Use this to retrieve the attachment files of an email. The attachment files are returned as the file IDs on Google Drive.
• get_current_date_time: The current date and time are returned. Please provide your timezone. If you don't know the timezone, it is automatically detected with the script.
• get_current_weather: Use this to get the current weather using the latitude and the longitude. At that time, convert the location to the latitude and the longitude and provide them to the function. If you cannot know the location, decide the location using the timezone.
• get_exchange_rate: Use this to get the current exchange rate. Using this, for example, it can exchange yen for dollars.
• get_file_from_google_drive: Use this to get and download a file from Google Drive by giving a filename. When you use this function, the returned data is base64 data. So, you are required to decode base64 data.
• get_google_doc_object_using_docs_api: Use this to get Google Docs Object using Docs API. https://developers.google.com/workspace/docs/api/reference/rest/v1/documents/get When this tool is used, for example, the index of each content in the document body can be retrieved. This cannot be directly used for retrieving text of the document body.
• get_google_sheet_object_using_sheets_api: Use this to get Google Sheets Object using Sheets API. https://developers.google.com/workspace/sheets/api/reference/rest/v4/spreadsheets/get When this tool is used, for example, the sheet names can be converted to sheet IDs. This cannot be used for retrieving the cell values.
• get_google_slides_object_using_slides_api: Use this to get Google Slides Object using Slides API. https://developers.google.com/workspace/slides/api/reference/rest/v1/presentations/get When this tool is used, for example, the object IDs on the slides can be retrieved.
• get_massages_by_search_from_Gmail: Get messages (emails) from Gmail using the search query. This function returns the messages using the search query.
• get_massages_by_time_from_Gmail: Get messages (emails) from Gmail using the time. This function returns the messages from "after" to now.
• get_specific_date_weather: Use this to get the weather for the specific date using the latitude and the longitude. At that time, convert the location to the latitude and the longitude and provide them to the function. The date is required to be included. The date format is "yyyy-MM-dd HH:mm" If you cannot know the location, decide the location using the timezone.
• get_values_from_google_docs: Use this to get text from Google Docs in a text format or a markdown format. The document ID is used for retrieving the values from the Google Docs. If you use the document URL, get the document ID from the URL and use the ID.
• get_values_from_google_sheets: Use this to get cell values from Google Sheets. The spreadsheet ID is used for retrieving the values from the Google Sheets. If you use the spreadsheet URL, get the spreadsheet ID from the URL and use the ID.
• manage_google_docs_using_docs_api: Use this to manage Google Docs using Docs API. Provide the request body for batchUpdate method. In order to retrieve the detailed information of the document, including the index and so on, it is required to use a tool "get_google_doc_object_using_docs_api".
• manage_google_sheets_using_sheets_api: Use this to manage Google Sheets using Sheets API. Provide the request body for batchUpdate method. In order to retrieve the detailed information of the spreadsheet, including the sheet ID and so on, it is required to use a tool "get_google_sheet_object_using_sheets_api".
• manage_google_slides_using_slides_api: Use this to manage Google Slides using Slides API. Provide the request body for batchUpdate method. In order to retrieve the detailed information of the spreadsheet, including the object ID and so on, it is required to use a tool "get_google_slides_object_using_slides_api".
• move_files_on_google_drive: Use this to move the files and the folders into other folder on Google Drive.
• people_connections_list: Provides a list of the authenticated user's contacts.
• people_contactGroups_list: List all contact groups owned by the authenticated user. Members of the contact groups are not populated.
• people_otherContacts_list: List all "Other contacts", that is contacts that are not in a contact group. "Other contacts" are typically auto created contacts from interactions.
• people_otherContacts_search: Use to provide a list of contacts in the authenticated user's other contacts that matches the search query. The query matches on a contact's names, emailAddresses, and phoneNumbers fields that are from the OTHER_CONTACT source.
• people_people_getBatchGet: Use to provide information about a list of specific people by specifying a list of requested resource names. Use people/me to indicate the authenticated user.
• put_file_to_google_drive: Use this to put and upload data to Google Drive as a file. When you use this function, please provide the file content converted to base64 data. So, you are required to encode the file content as base64 data.
• put_values_into_google_docs: Use this to append or insert text to Google Docs. The document ID is used for putting the values to the Google Docs. If you use the document URL, get the document ID from the URL and use the ID.
• put_values_to_google_sheets: Use this to put values into Google Sheets. The spreadsheet ID is used for putting the values to the Google Sheets. If you use the spreadsheet URL, get the spreadsheet ID from the URL and use the ID.
• remove_mails_Gmail: Use this to remove the messages.
• rename_files_on_google_drive: Use this to rename the files on Google Drive.
• revisions_drive_api_list: Use to get a list of a file's revisions on Google Drive.
• search_file_in_google_drive: Use this to search files in Google Drive by providing a search query. For example, the filename can be converted to the file ID. But, in the case of Google Drive, the file IDs are unique values. But, the same filenames can exist in the same folder. So, when a filename is searched, multiple file IDs might be returned. At that time, it is required to confirm which file the user wants to use.
• search_schedule_on_Google_Calendar: Use to search the schedules (events) on Google Calendar by providing the date range.
• search_values_from_google_sheets: Use this to search all cells in Google Sheets using a regex. The spreadsheet ID is used for searching a text from the Google Sheets. If you use the spreadsheet URL, get the spreadsheet ID from the URL and use the ID. In this case, the search text is searched to see whether it is the same as the entire cell value. So, if you want to search the cells including 'sample' text, please use a regex like '.sample.'.
• send_mails_Gmail: Use this to send the draft emails which have already been created. If you want to send an email, first, it is required to create a draft email. By this, the draft email can be sent.
• summarize_file_on_google_drive: Use this to describe and summaize a file on Google Drive.
• update_schedule_on_Google_Calendar: Use to update the schedule (event) on Google Calendar.

This section details deploying the MCP server using Google Apps Script and configuring the Gemini CLI to use it.

First, create a new standalone Google Apps Script project. A standalone project is not bound to a specific Google Sheet, Doc, or Form, making it ideal for creating a general-purpose web service. You can create one by visiting script.google.com. Ref

To simplify building the MCP server, we will use pre-built Google Apps Script libraries. These encapsulate the complex MCP handling logic and provide ready-to-use tools, keeping the main script clean.

This sample uses two Google Apps Script libraries:

1. MCPApp: Manages the MCP server lifecycle and communication protocols.
2. ToolsForMCPServer: Provides a suite of pre-built tools for interacting with Google Workspace services (Gmail, Drive, Calendar, etc.).

1. Open the script editor of the project you just created.
2. Install MCPApp:
   • Installation Guide
   • Project Key: 1TlX_L9COAriBlAYvrMLiRFQ5WVf1n0jChB6zHamq2TNwuSbVlI5sBUzh
   • Identifier: MCPApp
   • Repository: https://github.com/tanaikech/MCPApp
3. Install ToolsForMCPServer:
   • Installation Guide
   • Project Key: 1lnE7UL1jQgPDbTB9yjhiwZM0SaS9MObhzvWUWb_t8FisO6A3bLepvM2j
   • Identifier: ToolsForMCPServer
   • Repository: https://github.com/tanaikech/ToolsForMCPServer

Please copy and paste the following script into the script editor (replacing any existing code) and save the project.

This is a basic script for using this library.

const apiKey = "###"; // API key for Gemini API

/**
 * This function is automatically run when the MCP client accesses Web Apps.
 */
const doPost = e => main(e);

function main(eventObject) {
  const m = ToolsForMCPServer;

  m.apiKey = apiKey; // This is an API key for using Gemini API.
  // m.defaultCalendarId = "###"; // If you want to use the specific calendar, please use this.
  // m.enableBaseTools = false; // Disable base tools
  // m.enableClassroomTools = false; // Disable tools for managing Google Classroom
  // m.enablePeopleTools = false; // Disable tools for managing Google People
  // m.enableAnalyticsTools = false; // Disable tools for managing Google Analytics

  const object = { eventObject, items: m.getTools() };
  return new MCPApp
    .mcpApp({ accessKey: "sample" })
    .setServices({ lock: LockService.getScriptLock() })
    .server(object);
}

Note:

• If you intend to use the following tools, you must uncomment the apiKey line in the script and provide a valid API key for the Gemini API.
  • generate_roadmap_to_google_sheets: Creates a roadmap in Google Sheets.
  • generate_description_on_google_drive: Generates and sets a description for a file on Google Drive.
  • generate_image_on_google_drive: Generates an image from a prompt and saves it to Google Drive.
  • summarize_file_on_google_drive: Summarizes a file stored on Google Drive.
  • description_web_site: Provides descriptions of websites given their URLs.
• If an error related to Drive API occurred, please enable Drive API at Advanced Google services.
• If you want to manage Docs, Sheets, Slides, and Calendars using the batch update methods of API, please enable Docs API, Sheets API, Slides API, and Calendar API at Advanced Google services.
• If you want to use the specific Google Calendar, please set defaultCalendarId.
• If you want to manage Google Classroom and Google People, please enable the Google Classroom API and the Google People API at Advanced Google services.
• If you want to retrieve the drive activity, please enable Drive Activity API at Advanced Google services.
• If you want to use Google Analytics, please enable Google Analytics Admin API and Google Analytics Data API at Advanced Google services.

When this script is run, all tools in this library are shown as a JSON object.

function showAlltools() {
  const res = ToolsForMCPServer.getToolList();
  console.log(res);
}

When you want to use the specific tools, you can also use the following script.

This script uses only the tool get_exchange_rate.

function main(eventObject) {
  const enables = ["get_exchange_rate"];

  const m = ToolsForMCPServer;
  m.apiKey = apiKey;
  const object = { eventObject, items: m.getTools({ enables }) };
  return new MCPApp
    .mcpApp({ accessKey: "sample" })
    .setServices({ lock: LockService.getScriptLock() })
    .server(object);
}

This script uses all tools except for the tool get_exchange_rate.

function main(eventObject) {
  const disables = ["get_exchange_rate"];

  const m = ToolsForMCPServer;
  m.apiKey = apiKey;
  const object = { eventObject, items: m.getTools({ disables }) };
  return new MCPApp
    .mcpApp({ accessKey: "sample" })
    .setServices({ lock: LockService.getScriptLock() })
    .server(object);
}

To allow the Gemini CLI to communicate with our script, we must deploy it as a Web App. This creates a unique URL that acts as our MCP server endpoint.

You can find detailed information in the official documentation.

Please follow these steps to deploy the Web App in the script editor:

1. In the script editor, at the top right, click Deploy -> New deployment.
2. Click Select type -> Web App.
3. Enter a description for the Web App in the fields under Deployment configuration.
4. Select "Me" for "Execute as". This is crucial, as it allows the script to run with your permissions to access your Google services.
5. Select "Anyone" for "Who has access". This makes the URL callable from the internet. Access is controlled by the unguessable URL and the accessKey defined in the script.
6. Click Deploy.
7. After authorizing the necessary scopes, copy the Web app URL. It will look similar to https://script.google.com/macros/s/###/exec. This is your MCP server endpoint.

Important: When you modify the Apps Script code, you must create a new deployment version to publish the changes. Click Deploy > Manage deployments, select your active deployment, click the pencil icon, and choose "New version" from the Version dropdown. More info here.

Follow the official documentation to install the Gemini CLI on your system. Ref

To connect the Gemini CLI to your new Apps Script server, you need to edit its settings file. This file is typically located at ~/.gemini/settings.json on macOS/Linux or %USERPROFILE%\.gemini\settings.json on Windows.

Add the mcpServers configuration block as shown below.

1. Replace https://script.google.com/macros/s/###/exec with the Web App URL you copied earlier.
2. Ensure the accessKey query parameter matches the accessKey you defined in your Google Apps Script (sample in this example).

{
  "theme": "Default",
  "selectedAuthType": "gemini-api-key",
  "mcpServers": {
    "gas_web_apps": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://script.google.com/macros/s/###/exec?accessKey=sample"
      ],
      "env": {}
    }
  },
  "disableAutoUpdate": true
}

or

{
  "theme": "Default",
  "selectedAuthType": "###",
  "mcpServers": {
    "gas_web_apps": {
      "httpUrl": "https://script.google.com/macros/s/###/exec?accessKey=sample",
      "description": "MCP server built by Google Apps Script Web Apps"
    }
  }
}

• "gas_web_apps": A local identifier for your server.

• "httpUrl": Your Web Apps URL. In the current stage, when this is used, an error might occur. At that time, use mcp-remote.

• If you use mcp-remote, please install it. Ref

• If an error like Request timed out occurs, please add timeout and use GEMINI_TIMEOUT as follows.

  {
    "theme": "Default",
    "selectedAuthType": "gemini-api-key",
    "mcpServers": {
      "gas_web_apps": {
        "command": "npx",
        "args": [
          "mcp-remote",
          "https://script.google.com/macros/s/###/exec?accessKey=sample"
        ],
        "env": {},
        "timeout": 300000
      }
    },
    "disableAutoUpdate": true
  }

  GEMINI_TIMEOUT=300 # Seconds
  

Recently, I published a report titled "Processing File Content Using Gemini CLI with an MCP Server Built by Google Apps Script" Ref. In this report, I concluded that using the Drive API in the background is more effective for transferring file content between Gemini CLI and Google Drive than embedding base64 data directly in the prompt. Therefore, to facilitate this effective transfer, I use ggsrun, a CLI tool.

To use this method, you first need to set up ggsrun.

1. Download ggsrun: Get the latest release from the ggsrun releases page. The main repository is here.
2. Create a Google Cloud Project: Go to the Google Cloud Platform Resource Manager and click CREATE PROJECT.
3. Enable Drive API: In your new project, navigate to APIs & Services > Enabled APIs & Services. Search for and enable the Drive API.
4. Create Credentials: Navigate to APIs & Services > Credentials.
5. Configure OAuth Client ID: Click Create credentials, select OAuth client ID, and choose Web Application.
6. Set Redirect URI: Under "Authorized redirect URIs," add http://localhost:8080.
7. Create and Download Credentials: Click Create, then Download JSON. Important: Rename the downloaded file to client_secret.json and place it in your working directory.
8. Authenticate ggsrun:

8-1. Open your terminal in the directory containing client_secret.json.

8-2. Run the command ggsrun auth.

8-3. Copy the URL displayed in the terminal, paste it into your browser, and authorize the requested scopes.

8-4. Copy the authorization code from your browser's address bar and paste it back into the terminal.

To verify the setup, run ggsrun di to display information about your Google Drive.

To enable the Gemini CLI to use ggsrun, add the following tool definitions to your GEMINI.md file. This file is typically located at ~/.gemini/GEMINI.md on macOS/Linux or %USERPROFILE%\.gemini\GEMINI.md on Windows.

# Transferring file content between local PC and Google Drive using ggsrun

* GitHub Repository: [https://github.com/tanaikech/ggsrun](https://github.com/tanaikech/ggsrun)
* To download a file from Google Drive by filename, use this command:
  ```bash
  ggsrun d -f [filename]
  ```
* To download a file from Google Drive by file ID, use this command:
  ```bash
  ggsrun d -i [fileId]
  ```
* To download a file from Google Drive by a file ID by converting mimeType, use this command:
  ```bash
  ggsrun d -i [fileId] -e [extension]
  ```
  [extension] is the file extension. For example, when you want to download the file in PDF format and TEXT format, use "pdf" and "txt", respectively.
* To upload files from the local PC to the root directory of Google Drive, use this command:
  ```bash
  ggsrun u -f "[filename1],[filename2],,,"
  ```
* To upload files from the local PC to a specific directory of Google Drive, use this command:
  ```bash
  ggsrun u -p [folderId] -f "[filename1],[filename2],,,"
  ```
* To search for files and folders on Google Drive, use this command:
  ```bash
  ggsrun sf -q "[search query of Drive API v3 ([https://developers.google.com/workspace/drive/api/guides/search-shareddrives](https://developers.google.com/workspace/drive/api/guides/search-shareddrives))]"
  ```
* To search for files and folders on Google Drive, use this command:
  ```bash
  ggsrun ls -s [filename]
  ```

The following examples demonstrate how the Gemini CLI uses the Google Apps Script MCP server to interact with various Google services.

This command lists all configured MCP servers and the tools they provide.

Prompt

> /mcp

When your setting is correct, the following result is returned.

Answer

ℹConfigured MCP servers:

  🟢 gas_web_apps - Ready (125 tools, 3 prompts)
    Tools:
    - add_label_to_Gmail
    - analytics_admin_accountSummaries_list
    - analytics_admin_properties_get
    - analytics_data_properties_runRealtimeReport
    - analytics_data_properties_runReport
    - auto_new_draft_creation_Gmail
    - auto_reply_draft_creation_Gmail
    - change_permission_of_file_on_google_drive
    - classroom_courses_aliases_create
    - classroom_courses_aliases_delete
    - classroom_courses_aliases_list
    - classroom_courses_announcements_create
    - classroom_courses_announcements_delete
    - classroom_courses_announcements_get
    - classroom_courses_announcements_list
    - classroom_courses_announcements_modifyAssignees
    - classroom_courses_announcements_patch
    - classroom_courses_courseWork_create
    - classroom_courses_courseWork_delete
    - classroom_courses_courseWork_get
    - classroom_courses_courseWork_list
    - classroom_courses_courseWork_modifyAssignees
    - classroom_courses_courseWork_patch
    - classroom_courses_courseWork_rubrics_create
    - classroom_courses_courseWork_rubrics_delete
    - classroom_courses_courseWork_rubrics_get
    - classroom_courses_courseWork_rubrics_list
    - classroom_courses_courseWork_rubrics_patch
    - classroom_courses_courseWork_studentSubmissions_get
    - classroom_courses_courseWork_studentSubmissions_list
    - classroom_courses_courseWork_studentSubmissions_patch
    - classroom_courses_courseWork_studentSubmissions_reclaim
    - classroom_courses_courseWork_studentSubmissions_return
    - classroom_courses_courseWork_studentSubmissions_turnIn
    - classroom_courses_courseWorkMaterials_create
    - classroom_courses_courseWorkMaterials_delete
    - classroom_courses_courseWorkMaterials_get
    - classroom_courses_courseWorkMaterials_list
    - classroom_courses_courseWorkMaterials_patch
    - classroom_courses_create
    - classroom_courses_get
    - classroom_courses_getGradingPeriodSettings
    - classroom_courses_list
    - classroom_courses_patch
    - classroom_courses_remove
    - classroom_courses_students_create
    - classroom_courses_students_delete
    - classroom_courses_students_get
    - classroom_courses_students_list
    - classroom_courses_teachers_create
    - classroom_courses_teachers_delete
    - classroom_courses_teachers_get
    - classroom_courses_teachers_list
    - classroom_courses_topics_create
    - classroom_courses_topics_delete
    - classroom_courses_topics_get
    - classroom_courses_topics_list
    - classroom_courses_topics_patch
    - classroom_courses_update
    - classroom_courses_updateGradingPeriodSettings
    - classroom_invitations_accept
    - classroom_invitations_create
    - classroom_invitations_get
    - classroom_invitations_list
    - classroom_invitations_remove
    - classroom_registrations_create
    - classroom_registrations_delete
    - classroom_userProfiles_get
    - classroom_userProfiles_guardianInvitations_create
    - classroom_userProfiles_guardianInvitations_get
    - classroom_userProfiles_guardianInvitations_list
    - classroom_userProfiles_guardianInvitations_patch
    - classroom_userProfiles_guardians_get
    - classroom_userProfiles_guardians_list
    - classroom_userProfiles_guardians_remove
    - comments_drive_api_list
    - comments_drive_api_remove
    - convert_mimetype_of_file_on_google_drive
    - create_file_to_google_drive
    - create_google_docs_from_markdown_on_google_drive
    - create_schedule_on_Google_Calendar
    - delete_schedules_on_Google_Calendar
    - description_video_on_youtube
    - description_web_site
    - drive_activity_api_query
    - generate_description_on_google_drive
    - generate_image_on_google_drive
    - generate_presentation_with_google_slides
    - generate_quiz_with_google_forms
    - generate_roadmap_to_google_sheets
    - generate_survey_with_google_forms
    - get_attachment_files_from_Gmail
    - get_current_date_time
    - get_current_weather
    - get_exchange_rate
    - get_file_from_google_drive
    - get_google_doc_object_using_docs_api
    - get_google_sheet_object_using_sheets_api
    - get_google_slides_object_using_slides_api
    - get_massages_by_search_from_Gmail
    - get_massages_by_time_from_Gmail
    - get_specific_date_weather
    - get_values_from_google_docs
    - get_values_from_google_sheets
    - manage_google_docs_using_docs_api
    - manage_google_sheets_using_sheets_api
    - manage_google_slides_using_slides_api
    - move_files_on_google_drive
    - people_connections_list
    - people_contactGroups_list
    - people_otherContacts_list
    - people_otherContacts_search
    - people_people_getBatchGet
    - put_file_to_google_drive
    - put_values_into_google_docs
    - put_values_to_google_sheets
    - remove_mails_Gmail
    - rename_files_on_google_drive
    - revisions_drive_api_list
    - search_file_in_google_drive
    - search_schedule_on_Google_Calendar
    - search_values_from_google_sheets
    - send_mails_Gmail
    - summarize_file_on_google_drive
    - update_schedule_on_Google_Calendar

    Prompts:
    - generate_roadmap
    - get_weather
    - search_files_on_google_drive

Prompt

Search for files with the filename of "sample file" from Google Drive.

Answer

  ✔  search_file_in_google_drive (gas_web_apps MCP Server) {"query":"name = 'sample file'"}

✦ I found a file named "sample file" on Google Drive. It's a Google Sheet, and here's its link:
  https://docs.google.com/spreadsheets/d/###/edit?usp=drivesdk

Gemini interpreted the request and selected the search_file_in_google_drive tool from the gas_web_apps MCP server. It constructed the appropriate Google Drive search query (name = 'sample file') and executed the search via Google Apps Script, returning the file metadata.

Prompt

Tell me my today's schedule.

Answer

  ✔  search_schedule_on_Google_Calendar (gas_web_apps MCP Server) {"start":"2025-07-03","end":"2025-07-03"}

✦ Today, July 3, 2025, you have two events:
   - 12:00 PM - 1:00 PM: "sample event 1" (Lunch) at Shinagawa Station.
   - 7:00 PM - 8:00 PM: "sample event 2" (Dinner) at Osaka Station.

The model used the search_schedule_on_Google_Calendar tool, automatically determining the correct start and end dates for "today". The MCP server fetched the events from Google Calendar, and Gemini summarized the results.

Prompt

Tell me my today's emails from tanaike@hotmail.com.

Answer

  ✔  get_messages_by_search_from_Gmail (gas_web_apps MCP Server) {"query":"from:tanaike@hotmail.com after:2025/07/03"}

✦ Here are your emails from tanaike@hotmail.com for today:

   1. From: Tanaike Kanshi <tanaike@hotmail.com>
      Subject: sample mail 2
      Body: This is a sample mail of No.2.

   2. From: Tanaike Kanshi <tanaike@hotmail.com>
      Subject: sample mail 1
      Body: This is a sample mail of No.1.

Gemini utilized the get_messages_by_search_from_Gmail tool, correctly formulating a Gmail search query (from:tanaike@hotmail.com after:2025/07/03) to meet the constraints of the prompt.

Prompt

I received a souvenir from my friend Tanaike. Create an email to thank him for it. Tanaike's email address is tanaike@hotmail.com.

Answer

  ✔  auto_new_draft_creation_Gmail (gas_web_apps MCP Server) {"obj":[{"body":"Dear Tanaike,\n\nThank you so much for the wonder…
  A draft email was successfully created. The draft ID is "r-###".

  ✔  send_mails_Gmail (gas_web_apps MCP Server) {"draftIds":["r-###"]}
  Message of "Thank you for the souvenir!" was correctly sent.

✦ I have already sent the email to Tanaike thanking him for the souvenir.

The following email was automatically sent to tanaike@hotmail.com:

Dear Tanaike,

Thank you so much for the wonderful souvenir! I really appreciate your
thoughtfulness.

Best regards,
Kanshi Tanaike

This prompt required two steps. First, Gemini used auto_new_draft_creation_Gmail to generate the email content and create a draft within Gmail via the MCP server. Second, it immediately used the send_mails_Gmail tool, passing the draft ID from the first step to send the email.

This example demonstrates complex orchestration involving web search and multiple MCP tools. This prompt is similar to one found in this reference: Ref

Prompt

Check tomorrow's weather conditions at 12 PM around Shinagawa Station and exchange 100 US Dollars for Japanese Yen. Under this condition, will we enjoy eating lunch outside for 1 hour at the retrieved location and that day? Finally, create it as a schedule in Google Calendar.

Answer

  ✔  GoogleSearch Searching the web for: "latitude and longitude of Shinagawa Station"
  Search results for "latitude and longitude of Shinagawa Station" returned.
  
  ✔  get_current_date_time (gas_web_apps MCP Server) {"timezone":"Asia/Tokyo"}
  The current date and time are "2025-07-03 11:04:42". Timezone is "Asia/Tokyo".

  ✔  get_exchange_rate (gas_web_apps MCP Server) {"currency_from":"USD","currency_to":"JPY"}
  The raw data from the API is {"amount":1.0,"base":"USD","date":"2025-07-02","rates":{"JPY":144.15}}. The detailed result is as follows. The currency rate at latest from "USD" to "JPY" is 144.15. 
 
  ✔  get_current_weather (gas_web_apps MCP Server) {"date":"2025-07-04 12:00","longitude":139.73917,"latitude":35.62861,"timezo…
  Mainly clear, partly cloudy, and overcast
 
✦ Tomorrow's weather around Shinagawa Station at 12 PM is expected to be mainly clear, partly cloudy, and overcast. The exchange rate for 100 US Dollars is 14415 Japanese Yen (1 USD = 144.15 JPY).

  Given the weather conditions, it might be enjoyable to eat lunch outside, but there's a chance of it being overcast.
 
  ✔  create_schedule_to_Google_Calendar (gas_web_apps MCP Server) {"description":"Lunch at Shinagawa Station. Weather: Mainly c…
  An event was created as Start: 2025-07-04 12:00:00, End: 2025-07-04 13:00:00, Title: Outdoor Lunch at Shinagawa Station, Description: Lunch at Shinagawa Station. Weather: Mainly clear, partly cloudy, and overcast. Exchange rate: 1 USD = 144.15 JPY.
 
✦ The schedule for an outdoor lunch at Shinagawa Station tomorrow, July 4, 2025, from 12:00 PM to 1:00 PM, has been created in your Google Calendar. The description includes the weather forecast (mainly clear, partly cloudy, and overcast) and the exchange rate (1 USD = 144.15 JPY).

The following event was automatically created in Google Calendar:

Date: July 4, 2025, from 12:00 PM to 1:00 PM
Title: Outdoor Lunch at Shinagawa Station
Description: Lunch at Shinagawa Station. Weather: Mainly clear, partly cloudy, and overcast. Exchange rate: 1 USD = 144.15 JPY.

This involved a complex chain of actions:

1. Gemini used the standard GoogleSearch to find the latitude and longitude of Shinagawa Station.
2. It used the MCP tool get_current_date_time to establish the date.
3. It used get_exchange_rate to find the USD to JPY conversion.
4. It used get_current_weather, inputting the coordinates found in step 1 and the date/time.
5. Gemini synthesized this information to answer the qualitative question about enjoying lunch.
6. Finally, it used create_schedule_to_Google_Calendar, embedding the retrieved weather and exchange rate information into the event description.

Prompt

Append the following text to the document "sample document 20250703".
<Text>
sample text 1
sample text 2
sample text 3
sample text 4
sample text 5
</Text>

Answer

  ✔  search_file_in_google_drive (gas_web_apps MCP Server) {"query":"sample document 20250703"}
  "isError": true

  ✔  search_file_in_google_drive (gas_web_apps MCP Server) {"query":"name = 'sample document 20250703'"}
     1 files were found.

  ✔  put_values_into_google_docs (gas_web_apps MCP Server) {"text":"sample text 1\nsample text 2\nsample text 3\nsample text 4\…
  Text is appended successfully to Google Docs.

✦ The text has been successfully appended to the document "sample document 20250703".

This example demonstrates Gemini's ability to self-correct when using tools.

1. Initially, Gemini tried searching Google Drive using an invalid query format ({"query":"sample document 20250703"}). The MCP server returned an error ("isError": true).
2. Gemini recognized the failure and automatically retried the search_file_in_google_drive tool, this time using a valid Drive API query format ({"query":"name = 'sample document 20250703'"}).
3. Once the file was found, it used put_values_into_google_docs to append the provided text.

Prompt

Create a new Google Document and put the following text into the Google Document.
<Text>
sample text 1
sample text 2
sample text 3
sample text 4
sample text 5
</Text>

Answer

  ✔  create_file_to_google_drive (gas_web_apps MCP Server) {"mimeType":"application/vnd.google-apps.document","filename":"New D…
  A file was created on the root folder. The file URL is "https://docs.google.com/document/d/###/edit?usp=drivesdk".

  ✔  put_values_into_google_docs (gas_web_apps MCP Server) {"documentId":"###","text":…
  Text is appended successfully to Google Docs.

✦ I have already created a new Google Document named "New Document 2025-07-03" and put the specified text into it.

This required two sequential tools. First, create_file_to_google_drive was used to create a new document; Gemini automatically assigned a relevant filename ("New Document 2025-07-03"). Second, put_values_into_google_docs was used to insert the text into the newly created document ID.

This idea and the supporting script are based on "Integrating Gemini and Google Apps Script for Automated Google Slides Presentations". Ref

Prompt

Please create a 3-minute presentation. The title will be "Cooking miso soup". My name is Tanaike. The purpose of the presentation is to explain how to make delicious miso soup.

Answer

  ✔  generate_presentation_with_google_slides (gas_web_apps MCP Server) {"name":"Tanaike","text":"This presentation will guide …
  Presentation was successfully created. The url is "https://docs.google.com/open?id=###".

✦ The presentation "Cooking misosoup" has been created. You can access it at:
  https://docs.google.com/open?id=###

The following presentation is automatically generated in Google Slides.

The Gemini CLI passed the request parameters to the generate_presentation_with_google_slides tool on the MCP server. This specific tool in Google Apps Script uses the Gemini API internally (requiring the API key mentioned in the setup) to generate the slide content and structure, then uses the Slides Service to create the presentation.

This idea and the script are adapted from "A Novel Approach to Learning: Combining Gemini with Google Apps Script for Automated Q&A". Ref

Prompt

I want to conduct a survey about favorite Japanese foods. Your mission is as follows:
* Generate a survey using Google Forms.
* Send an email to tanaike@hotmail.com including the published URL of the Google Form.
* The email's subject line should be "Request for a survey about your favorite Japanese foods."
* Create the email's body from the subject.

Answer

  ✔  generate_survey_with_google_forms (gas_web_apps MCP Server) {"title":"Favorite Japanese Foods Survey","itemList":[{"requir… 
  The survey was successfully generated as a Google Form. The edit URL and published URL of the form are "https://docs.google.com/forms/d/###/edit" and "https://docs.google.com/forms/d/e/#####/viewform", respectively.

  ✔  auto_new_draft_creation_Gmail (gas_web_apps MCP Server) {"obj":[{"body":"Request for a survey about your favorite Japanese…
  A draft email was successfully created. The draft ID is "r###".                                    

  ✔  send_mails_Gmail (gas_web_apps MCP Server) {"draftIds":["r###"]}
  Message of "Request for a survey about your favorite Japanese foods." was correctly sent.

✦ I have successfully generated the survey and sent the email with the published URL of the Google Form to tanaike@hotmail.com.

The following survey is automatically generated in Google Forms.

And the following email is sent to tanaike@hotmail.com:

Mail title:
Request for a survey about your favorite Japanese foods.

Mail body:
Request for a survey about your favorite Japanese foods.
Here is the link to the survey: https://docs.google.com/forms/d/e/#####/viewform

This required three steps integrating Forms and Gmail:

1. generate_survey_with_google_forms created the form and returned the published URL.
2. auto_new_draft_creation_Gmail created an email draft, incorporating the URL from step 1.
3. send_mails_Gmail sent the drafted email.

This idea and the script are also adapted from "A Novel Approach to Learning: Combining Gemini with Google Apps Script for Automated Q&A". Ref

Prompt

Suggest 5 questions and answers for a quiz with multiple-choice questions for learning Google Apps Script for beginners. Generate the quiz in Google Forms using the questions. The number of choices for each question is 5. The title of the quiz is "Learning Google Apps Script".

Answer

  ✔  generate_quiz_with_google_forms (gas_web_apps MCP Server) {"title":"Learning Google Apps Script","itemList":[{"required":t… 
  The quiz was successfully generated as a Google Form. The edit URL and published URL of the form are "https://docs.google.com/forms/d/###/edit" and "https://docs.google.com/forms/d/e/#####/viewform", respectively.

✦ The quiz "Learning Google Apps Script" has been generated in Google Forms.
   * Edit URL: https://docs.google.com/forms/d/###/edit
   * Published URL: https://docs.google.com/forms/d/e/#####/viewform

The following quiz is automatically generated in Google Forms.

Gemini generated the required questions, choices, and correct answers based on the prompt's requirements, and then structured this data into the parameters required by the generate_quiz_with_google_forms tool, which created the Google Form.

The following samples were added on July 9, 2025.

This example downloads an image file from Google Drive using the ggsrun CLI tool.

Prompt

Download the file "sample.png" from Google Drive using ggsrun.

Answer

  ✔  Shell ggsrun d -f sample.png (Download "sample.png" from Google Drive.)
     {"id":"###","name":"sample.png",,,

✦ The file "sample.png" has been downloaded.

When this prompt is run, the file sample.png is downloaded from Google Drive using the ggsrun CLI tool.

This example downloads an image file from a website using a shell command and uploads it to Google Drive using the ggsrun CLI tool.

Prompt The image at https://www.gstatic.com/marketing-cms/assets/images/7e/a4/253561a944f4a8f5e6dec4f5f26f/gemini.webp=s96-fcrop64=1,00000000ffffffff-rw is the Gemini icon from the official brand resource center. Ref

Download an image from the URL "https://www.gstatic.com/marketing-cms/assets/images/7e/a4/253561a944f4a8f5e6dec4f5f26f/gemini.webp=s96-fcrop64=1,00000000ffffffff-rw" and upload it to Google Drive using ggsrun.

Answer

  ✔  Shell wget -O /home/adsam/temp/workspace/gemini.webp "https://www.gstatic.com/marketing-cms/assets/images/7e/a4/253561a944… 

  ✔  Shell ggsrun u -f "gemini.webp" (Upload the downloaded image to Google Drive using ggsrun.)
     Error: Extension of '.webp' cannot be uploaded.

  ✔  Shell convert /home/adsam/temp/workspace/gemini.webp /home/adsam/temp/workspace/gemini.png (Convert the downloaded .webp image...) 

  ✔  Shell ggsrun u -f "gemini.png" (Upload the converted .png image to Google Drive using ggsrun.)
     {"upload_file_name":["gemini.png"],"uploaded_files":[{"id":"###","name":"gemini.png","mimeType":"image/png",,,

✦ The image has been successfully downloaded from the URL, converted to PNG format, and uploaded to Google Drive.

When this prompt is run, the following steps are executed:

1. An image file is downloaded using wget.
2. An attempt to upload the .webp file with ggsrun fails.
3. The file is converted to the PNG format using the convert command.
4. The converted .png file is successfully uploaded to Google Drive.

This example uploads a file to Google Drive using the ggsrun CLI tool and renames it using the rename_files_on_google_drive tool from the MCP server.

Prompt

Upload the file "sample.png" to Google Drive using ggsrun with the filename "test.png". And, show the file ID of the uploaded file.

Answer

  ✔  Shell ggsrun u -f "sample.png:test.png" (Attempt to upload "sample.png" to Google Drive as "test.png"...) 
     Error: open sample.png:test.png: no such file or directory.

  ✔  Shell ggsrun u -f "sample.png" (Upload "sample.png" to Google Drive.)
     {"upload_file_name":["sample.png"],"uploaded_files":[{"id":"###","name":"sample.png",,,

  ✔  rename_files_on_google_drive (gas_web_apps MCP Server) {"fileList":[{"fileId":"###","newName… 
     FileId: "###". Renamed successfully from "sample.png" to "test.png".

 ✦ The file "sample.png" has been uploaded to Google Drive and renamed to "test.png". The file ID of the uploaded file is ###.

When this prompt is run, the following steps are executed:

1. Upload a file from the local PC to Google Drive using ggsrun.
2. Rename the uploaded file using the rename_files_on_google_drive tool from the MCP server.
3. Return the file ID of the uploaded file.

Prompt

Upload the file "sample.png" using ggsrun, create a draft email with the uploaded file as an attachment, and send the email to "tanaike@hotmail.com".

Answer

  ✔  Shell ggsrun u -f "sample.png" (Uploads "sample.png" to Google Drive.)
     {"upload_file_name":["sample.png"],"uploaded_files":[{"id":"###","name":"sample.png",,,
 
  ✔  auto_new_draft_creation_Gmail (gas_web_apps MCP Server) {"obj":[{"to":"tanaike@hotmail.com","attachmentFiles":["###… 
     A draft email was successfully created. The draft ID is "r###".

   ✔  send_mails_Gmail (gas_web_apps MCP Server) {"draftIds":["r###"]}
     Message of "Uploaded File: sample.png" was correctly sent.

✦ I have uploaded the file "sample.png", created a draft email with the uploaded file as an attachment, and sent the email to "tanaike@hotmail.com".

A file is uploaded to Google Drive using ggsrun. Then, a new draft email is created with the uploaded file as an attachment using the auto_new_draft_creation_Gmail tool. Finally, the draft email is sent using the send_mails_Gmail tool.

Prompt

Move the file "sample.png" to the folder "temp_folder" on Google Drive.

Answer

 ✔  search_file_in_google_drive (gas_web_apps MCP Server) {"query":"name = 'sample.png'"}
    1 files were found.
    The list of found files is returned in the "FileList" JSON array.

 ✔  search_file_in_google_drive (gas_web_apps MCP Server) {"query":"name = 'temp_folder' and mimeType = 'application/vnd.googl…
    1 files were found.
    The list of found files is returned in the "FileList" JSON array.

 ✔  move_files_on_google_drive (gas_web_apps MCP Server) {"fileList":[{"dstId":"###","srcId":"1j…
    File "sample.png" was moved to the folder "temp_folder".

✦ I have moved the file "sample.png" to the folder "temp_folder" on Google Drive.

This prompt executes the following steps:

1. The file and folder are located by name to retrieve their IDs using the search_file_in_google_drive tool.
2. The file is moved to the destination folder using the move_files_on_google_drive tool.

The sample PDF file is PhysRev.48.73.pdf. This is from this page.

Prompt

Upload the PDF file PhysRev.48.73.pdf to Google Drive using ggsrun and set the description for the file.

Answer

  ✔  Shell ggsrun u -f "PhysRev.48.73.pdf" (Uploads the PDF file "PhysRev.48.73.pdf" to Google Drive.)
     {"upload_file_name":["PhysRev.48.73.pdf"],"uploaded_files":[{"id":"###",,,

  ✔  generate_description_on_google_drive (gas_web_apps MCP Server) {"fileId":"###"}
     The created description is "This file is a scientific paper titled "The Particle Problem in the General Theory of Relativity" authored by A. Einstein and N. Rosen, published in Physical Review, Volume 48, pages 73-77, on July 1, 1935...

✦ The PDF file PhysRev.48.73.pdf has been uploaded to Google Drive and its description has been set.

In this prompt, the PDF file PhysRev.48.73.pdf is uploaded to Google Drive using ggsrun. Then, a description for the PDF file is generated and set using the generate_description_on_google_drive tool from the MCP server.

Prompt

Generate an image of orange juice being poured into a glass, and download it.

Answer

  ✔  generate_image_on_google_drive (gas_web_apps MCP Server) {"prompt":"orange juice is poured into a glass"}
     An image was successfully generated from a prompt "orange juice is poured into a glass" as a file on Google Drive. The file URL and ID are "https://drive.google.com/file/d/###/view?usp=drivesdk" and "###", respectively.
     
  ✔  Shell ggsrun d -i ### -o "orange juice is poured into a glass" (Downloads the generated image...) 
     {"id":"###","name":"orange juice is poured into a glass",,,,

✦ The image "orange juice is poured into a glass" has been generated and downloaded to your current directory.

The following image was generated and downloaded.

An image is generated using the generate_image_on_google_drive tool from the MCP server and then downloaded using ggsrun.

Prompt

Describe the website "https://tanaikech.github.io/about/".

Answer

  ✔  description_web_site (gas_web_apps MCP Server) {"urls":["https://tanaikech.github.io/about/"]}
     The website "About · tanaike" belongs to a Japanese scientist with a Ph.D. in Physics, who is also a Google Developer Expert (GDE) in Google Workspace and a Google Cloud Champion Innovator. The individual focuses on creating novel solutions, with some inventions finding practical applications in various industries such as electronics, industrial machinery, architecture, and aerospace. The site provides contact information, including a developer profile, and links to their posts on the Google Cloud Blog, Medium, and Zenn. It also lists their activities on platforms like GitHub, Gists, Qiita, Stack Overflow, and X, showcasing applications, tips, and summaries of Google Apps Script related repositories. The site also includes information about donations via PayPal.Me.

✦ The website "About · tanaike" describes a Japanese scientist with a Ph.D. in Physics, who is also a Google Developer Expert (GDE) in Google Workspace and a Google Cloud Champion Innovator. The individual focuses on creating novel solutions, with some inventions finding practical applications in various industries. The site provides contact information and links to their posts on various platforms like Google Cloud Blog, Medium, Zenn, GitHub, and Stack Overflow, showcasing applications, tips, and summaries of Google Apps Script related repositories.

In this example, a description of the website is generated using the description_web_site tool from the MCP server.

Prompt

Upload the PDF file PhysRev.48.73.pdf to Google Drive using ggsrun and return a summary of the PDF file within 100 words.

Answer

  ✔  Shell ggsrun u -f "PhysRev.48.73.pdf" (Uploads the PDF file to Google Drive.)
     {"upload_file_name":["PhysRev.48.73.pdf"],"uploaded_files":[{"id":"###",,,

  ✔  summarize_file_on_google_drive (gas_web_apps MCP Server) {"prompt":"Summarize this PDF file within 100 words.","fileId":"1…
     Einstein and Rosen investigate the particle problem in general relativity, aiming to describe elementary particles without singularities. They propose a slight modification to the gravitational equations, allowing for regular solutions. In this framework, particles are represented as "bridges" connecting two identical "sheets" of space. This approach provides a singularity-free description for neutral and charged particles. The most natural charged particle in this model is found to have zero gravitational mass. The theory offers a unified, consistent view using only gravitational and electromagnetic fields, explaining the atomistic nature of matter and the absence of negative neutral mass.

✦ The PDF file "PhysRev.48.73.pdf" has been uploaded to Google Drive and summarized as follows:
  "Einstein and Rosen investigate the particle problem in general relativity, aiming to describe elementary particles without singularities. They propose a slight modification to the gravitational equations, allowing for regular solutions. In this framework, particles are represented as 'bridges' connecting two identical 'sheets' of space. This approach provides a singularity-free description for neutral and charged particles. The most natural charged particle in this model is found to have zero gravitational mass. The theory offers a unified, consistent view using only gravitational and electromagnetic fields, explaining the atomistic nature of matter and the absence of negative neutral mass."

A PDF file is uploaded to Google Drive using ggsrun, and a summary is generated using the summarize_file_on_google_drive tool from the MCP server.

Prompt

I want to cook miso soup. To achieve this goal, create a new Google Spreadsheet, generate a roadmap for cooking miso soup in the spreadsheet, and download it as a PDF file using ggsrun.

Answer

  ✔  create_file_to_google_drive (gas_web_apps MCP Server) {"filename":"Miso Soup Roadmap","mimeType":"application/vnd.google-a…
     A file was created on the root folder. The file URL is "https://docs.google.com/spreadsheets/d/###/edit?usp=drivesdk".

  ✔  generate_roadmap_to_google_sheets (gas_web_apps MCP Server) {"spreadsheetId":"###…
     Roadmap was successfully generated in Google Sheets. The spreadsheet ID and the sheet name are "###" and "Sheet1" (sheet ID: "0"), respectively. The URL is "https://docs.google.com/spreadsheets/d/###/edit". The file ID of the converted PDF file is "###".

✦ I can run ggsrun d -i ###. This will download the PDF file of the miso soup roadmap from Google Drive to your local machine.

  ✔  Shell ggsrun d -i ### (Download the generated PDF roadmap from Google Drive.)

✦ I have successfully created the Google Spreadsheet, generated the roadmap for cooking miso soup, and downloaded it as a PDF file named Cook Miso Soup.pdf.

As a result, the following roadmap is generated and saved as a PDF file:

When this prompt is run, the following steps are executed:

1. Create a new Google Spreadsheet using the create_file_to_google_drive tool from the MCP server.
2. Generate a roadmap in the new spreadsheet using the generate_roadmap_to_google_sheets tool.
3. Download the roadmap as a PDF file from Google Drive using ggsrun.

Prompt

Describe the Web site of "https://blog.google/technology/developers/introducing-gemini-cli-open-source-ai-agent/", and think of a title and create a new Google Document and put the generated description into the document by giving the title. And, generate a presentation of Google Slides using the title and document. And, create an email including the URL of the presentation and send it to "tanaike@hotmail.com".

Answer

  ✔  description_web_site (gas_web_apps MCP Server) {"urls":["https://blog.google/technology/developers/introducing-gemini-cli-…
     The site describes the introduction of Gemini CLI, an open-source AI agent that integrates Gemini directly into
     developers' terminals. This tool offers free access to Gemini 2.5 Pro for individual Google account holders,
     with high usage limits for model requests. Gemini CLI aims to enhance the command-line experience by enabling
     coding, problem-solving, content generation, and task management using natural language. It shares technology
     with Gemini Code Assist, providing AI assistance in both the terminal and VS Code. The article highlights
     Gemini CLI's open and extensible nature, encouraging community contributions.

  ✔  create_file_to_google_drive (gas_web_apps MCP Server) {"mimeType":"application/vnd.google-apps.document","filename":"Intro…
     A file was created on the root folder. The file URL is "https://docs.google.com/document/d/###/edit?usp=drivesdk".

  ✔  put_values_into_google_docs (gas_web_apps MCP Server) {"text":"The site describes the introduction of Gemini CLI, an open-…
     Text is appended successfully to Google Docs.

  ✔  generate_presentation_with_google_slides (gas_web_apps MCP Server) {"name":"Kanshi Tanaike","title":"Introducing Gemini CL…
     Presentation was successfully created. The url is "https://docs.google.com/open?id=###".

  ✔  auto_new_draft_creation_Gmail (gas_web_apps MCP Server) {"obj":[{"body":"Here is the presentation on Introducing Gemini CL…
     A draft email was successfully created. The draft ID is "r-###".

  ✔  send_mails_Gmail (gas_web_apps MCP Server) {"draftIds":["r-###"]}
     Message of "Presentation: Introducing Gemini CLI: An Open-Source AI Agent" was correctly sent.
 
✦ I have successfully described the website, created a Google Document with the description and a title, generated a Google Slides presentation, and sent an email with the presentation URL to "tanaike@hotmail.com".

The following email was sent.

Mail subject: Presentation: Introducing Gemini CLI: An Open-Source AI Agent
Mail body: Here is the presentation on Introducing Gemini CLI: An Open-Source AI Agent: https://docs.google.com/open?id=###

When the URL is opened, the following presentation was generated.

When this prompt is run, the following steps are executed:

1. A tool description_web_site generates a description of the web site.
2. A tool create_file_to_google_drive creates a new Google Document.
3. A tool put_values_into_google_docs puts the generated description into the created document.
4. A tool generate_presentation_with_google_slides generates a presentation to Google Slides using the title and Google Document.
5. A tool auto_new_draft_creation_Gmail creates a draft email including the URL of the Google Slides.
6. A tool send_mails_Gmail sends the email.

The following samples were added on July 10, 2025.

Prompt

Create a new Google Document and send an email to tanaike@hotmail.com that includes the document. Grant the user edit permission for the Google Document.

Answer

  ✔  create_file_to_google_drive (gas_web_apps MCP Server) {"filename":"New Google Document","mimeType":"application/vnd.google…
     A file was created on the root folder. The file URL is "https://docs.google.com/document/d/###/edit?usp=drivesdk".
     
  ✔  change_permission_of_file_on_google_drive (gas_web_apps MCP Server) {"role":"editor","fileId":"###…
     Permission for the file 'New Google Document' (ID: ###) was successfully updated. User 'tanaike@hotmail.com' has been granted 'editor' access.
     
  ✔  auto_new_draft_creation_Gmail (gas_web_apps MCP Server) {"obj":[{"title":"New Google Document","to":"tanaike@hotmail.com",…
     A draft email was successfully created. The draft ID is "r###".
     
  ✔  send_mails_Gmail (gas_web_apps MCP Server) {"draftIds":["r###"]}
     Message of "New Google Document" was correctly sent.

✦ I have created a new Google Document, granted edit permission to tanaike@hotmail.com, and sent an email to them with the
  document link.

When this is run, an email including the URL of the Google Document. And, the write permission was added to the email.

The following samples were added on July 24, 2025.

Prompt

Use a tool "description_video_on_youtube". Describe a YouTube video by giving the URL "https://www.youtube.com/watch?v=Dc2WPQkovZE".

Answer

In this case, ggsrun of a CLI tool is used for exporting the file from Google Drive. So, GEMINI.md is set. Ref

Prompt

Please follow these steps:

1.  Review @PhysRev.48.73.pdf.
2.  Summarize the PDF content in Markdown format.
3.  Generate a Google Document from the Markdown as a temporary file.
4.  Export the Google Document as a PDF using ggsrun.
5.  Remove the temporary file.

Answer

By this flow, a PDF file is saved as a file in the working directory.

The following samples were added on July 29, 2025.

In this case, you can see the samples at "Next-Level Data Automation: Gemini CLI, Google Calendar, and MCP".

The following samples were added on August 1, 2025.

In this case, you can see the samples at "Enhanced Guide to Using Prompts in Gemini CLI".

The following samples were added on August 10, 2025.

In this case, you can see the samples at "Next-Level Classroom Automation: Gemini CLI, Google Classroom, and MCP".

The following samples were added on August 18, 2025.

In this case, you can see the samples at "Unifying Google Workspace with Natural Language: Integrated Collaboration through Gemini CLI and MCP".

The following samples were added on August 22, 2025.

In this case, you can see the samples at "Streamlining Web Page Insights with Natural Language using Gemini CLI, Google Analytics, and MCP".

The examples above demonstrate that combining the Gemini CLI with an MCP server built using Google Apps Script Web Apps enables powerful automation across Google Workspace. By leveraging Google Apps Script's inherent authorization capabilities, we can easily give Gemini access to Gmail, Calendar, Drive, Docs, Sheets, and Slides.

The samples provided represent only a small subset of what is possible. The ToolsForMCPServer library includes many functions derived from previously developed Google Apps Script utilities. Furthermore, because the MCP server is simply a Google Apps Script project, developers can easily incorporate their own existing GAS snippets as new tools for the Gemini CLI. As this MCP server is updated with more tools, the potential for complex, AI-driven automation within the Google ecosystem becomes nearly limitless.

• When Gemini CLI is launched and an error Error connecting to MCP server 'gas_web_apps': Connection failed for 'gas_web_apps': MCP error -32000: Connection closed occurs, please redeploy Web Apps to reflect the latest script and try it again.

You might want to add your custom tools and promts to the MCP server. At that time, please refer to the following script.

If you are required to add only tools, please remove prompts/list and prompts/get from itemsForMCP. If you are required to add only prompts, please remove tools/list from itemsForMCP.

/**
 * If you want to add your custom tools, please use the following script.
 * This is a simple sample. Please modify the following script to your situation.
 */
function getCustomTools() {
  const functions = {
    params_: {
      tool1: {
        description: "Use this for testing a tool 1 of MCP server.",
        parameters: {
          type: "object",
          properties: { sample: { type: "string", description: "Sample value." } },
          required: ["sample"]
        }
      },
      tool2: {
        description: "Use this for testing a tool 2 of MCP server.",
        parameters: {
          type: "object",
          properties: { sample: { type: "string", description: "Sample value." } },
          required: ["sample"]
        },
      }
    },

    tool1: (object) => object,
    tool2: (object) => object,
  };

  // for MCP
  const itemsForMCP = [
    ...Object.keys(functions.params_).map(f => (
      {
        "type": "tools/list",
        "function": functions[f],
        "value": {
          name: f,
          description: functions.params_[f].description,
          inputSchema: functions.params_[f].parameters,
        }
      })),

    {
      type: "prompts/list",
      value: {
        prompts: [
          {
            name: "prompt1",
            description: "Custom prompt 1",
            arguments: [
              { name: "sample1", description: "sample1", required: true },
            ],
          },
          {
            name: "prompt2",
            description: "Custom prompt 2",
            arguments: [
              { name: "sample2", description: "sample2", required: true },
            ],
          },
        ]
      }
    },

    {
      type: "prompts/get",
      value: {
        prompt1: {
          description: "Custom prompt 1",
          messages: [
            {
              role: "user",
              content: {
                type: "text",
                text: "Custom prompt 1",
              },
            },
          ],
        },
        prompt2: {
          description: "Custom prompt 2",
          messages: [
            {
              role: "user",
              content: {
                type: "text",
                text: "Custom prompt 2",
              },
            },
          ],
        },
      }
    },
  ];

  return itemsForMCP;
}

const apiKey = "###"; // API key for Gemini API

/**
 * This function is automatically run when the MCP client accesses Web Apps.
 */
const doPost = e => main(e);

function main(eventObject) {
  const m = ToolsForMCPServer;
  m.apiKey = apiKey;
  const object = { eventObject, items: [...m.getTools(), ...getCustomTools()] };
  return new MCPApp
    .mcpApp({ accessKey: "sample" })
    .setServices({ lock: LockService.getScriptLock() })
    .server(object);
}

MIT

Tanaike

Donate

• v1.0.0 (July 3, 2025)

  1. Initial release.

• v1.0.1 (July 9, 2025)

  1. The method name of getServer was changed to getTools. But, getServer can still be used.
  2. The following methods were added.
  • For management of Gmail
    • get_attachment_files_from_Gmail
  • For management of Google Drive
    • rename_files_on_google_drive
    • move_files_on_google_drive
    • convert_mimetype_of_file_on_google_drive
  • For using Gemini
    • generate_roadmap_to_google_sheets
    • generate_description_on_google_drive
    • generate_image_on_google_drive
    • summarize_file_on_google_drive
    • description_web_site
  3. Added sample prompts and answers in README.md.

• v1.0.2 (July 10, 2025)

  1. A tool change_permission_of_file_on_google_drive was added. By this, when a Google Docs files are included in Gmail, the email can be sent by giving permission to the Google Docs file.

• v1.0.3 (July 11, 2025)

  1. A tool get_current_weather was updated.
  2. A new tool get_specific_date_weather was added.

• v1.0.4 (July 11, 2025)

  1. A function getSheet_ was updated.

• v1.0.5 (July 16, 2025)

  1. A new tool search_values_from_google_sheets was added.

• v1.0.6 (July 17, 2025)

  1. The following 6 new tools were added.
  • get_google_doc_object_using_docs_api
  • manage_google_docs_using_docs_api
  • get_google_sheet_object_using_sheets_api
  • manage_google_sheets_using_sheets_api
  • get_google_slides_object_using_slides_api
  • manage_google_slides_using_slides_api

• v1.0.7 (July 19, 2025)

  1. Added a getToolList method for retrieving all current tools in the library.
  2. Tools can be filtered using enables or disables as an array argument for the getTools method. If enables is used, only the tools specified in the enables array will be used. If disables is used, all tools except those specified in the disables array will be used. If neither enables nor disables is used, all tools will be used.

• v1.0.8 (July 23, 2025)

  1. An issue occurred when I updated Gemini CLI from v0.1.12 to v0.1.13. Ref Fortunately, Google is already aware of this issue, and I'm awaiting a resolution. In the meantime, I've received emails about it, so I've updated ToolsForMCPServer for Gemini CLI v0.1.13. The detailed updates are as follows: I confirmed that all tools in ToolsForMCPServer v1.0.8 worked when tested with Gemini CLI v0.1.13.
  • oneOf has been removed from the schema of each tool.
  • Following this report, the request body is now generated on the MCP server side. Therefore, when using the tools manage_google_docs_using_docs_api, manage_google_sheets_using_sheets_api, and manage_google_slides_using_slides_api, please use your API key for the Gemini API.

• v1.0.9 (July 24, 2025)

  1. The following 2 new tools were added.
  • description_video_on_youtube: Describe a YouTube video by providing the URL.
  • create_google_docs_from_markdown_on_google_drive: Create a Google Document from a markdown format.

• v1.0.10 (July 26, 2025)

  1. When I updated Gemini CLI from v0.1.12 to v0.1.13, an issue related to the schema of MCP occurred. Ref So, as a workaround at the time, I updated this library. But when I updated Gemini CLI to v0.1.14, I confirmed that the previous schema could be used. So, I reimplemented the previous schema. By this, the request body for APIs can be directly generated using Gemini CLI v0.1.14.

• v1.0.11 (July 29, 2025)

  1. management_calendar.js was updated. search_schedule_on_Google_Calendar and create_schedule_to_Google_Calendar were updated. And, 2 tools delete_schedules_on_Google_Calendar and update_schedule_on_Google_Calendar were added. From this version, when you want to manage Google Calendars, please enable Calendar API at Advanced Google services.

• v1.0.12 (July 31, 2025)

  1. At Gemini CLI v0.1.15, prompts/list was called even when prompts wasn't included in capabilities. This resulted in the error Error discovering prompts from gas_web_apps: MCP error -32001: Request timed out when prompts wasn't returned for prompts/list. To resolve this, I updated ToolsForMCPServer to return an empty array for prompts, which eliminated the error. Consequently, with this update in v1.0.12, you can now set custom prompts and resources.

• v1.0.13 (August 1, 2025)

  1. prompts/get was updated. And, 3 prompts were added.

• v1.0.14 (August 2, 2025)

  1. The library version and protocol version were set as variables.

• v1.0.15 (August 2, 2025)

  1. Bugs of tools create_schedule_on_Google_Calendar and update_schedule_on_Google_Calendar were removed.

• v1.0.16 (August 5, 2025)

  1. A tool update_schedule_on_Google_Calendar was updated. From this version, when the event is updated, an email is sent. And, the guests can also be removed.

• v1.0.17 (August 5, 2025)

  1. main_library_side.js was updated.

• v1.0.18 (August 9, 2025)

  1. A bug was removed. When a tool for Google Forms is run, an error like FunctionDeclaration in the request exceeds the maximum allowed nesting depth. occurs. This bug was removed by refactoring the JSON schema.

• v1.0.19 (August 10, 2025)

  1. 67 Tools for managing Google Classroom and 3 tools for managing Google People were added. When you use those tools, please enable the Google Classroom API and the Google People API at Advanced Google services. You can see the sample prompts at Next-Level Classroom Automation: Gemini CLI, Google Classroom, and MCP.

• v1.0.20 (August 11, 2025)

  1. A bug was removed. Ref

• v1.0.21 (August 12, 2025)

  1. A tool create_schedule_on_Google_Calendar in management_calendar.js was updated. From this version, the Calendar API is used.

• v1.0.22 (August 13, 2025)

  1. Upon updating the Gemini CLI from v0.1.18 to v0.1.19, I noticed a regression. Ref I modified most of the tools, but left manage_google_docs_using_docs_api, manage_google_sheets_using_sheets_api, and manage_google_slides_using_slides_api untouched. These tools rely on $ref in their input schemas, but it appears $ref is not supported in version v0.1.19. I'll hold off on further changes and wait for a future update to the Gemini CLI. At v0.1.19, when Gemini CLI is run, the following 3 errors are shown in the above. But, other 112 tools can be used.

  Skipping tool 'manage_google_docs_using_docs_api' from MCP server 'gas_web_apps' because it has missing types in its parameter schema. Please file an issue with the owner of the MCP server.
  Skipping tool 'manage_google_sheets_using_sheets_api' from MCP server 'gas_web_apps' because it has missing types in its parameter schema. Please file an issue with the owner of the MCP server.
  Skipping tool 'manage_google_slides_using_slides_api' from MCP server 'gas_web_apps' because it has missing types in its parameter schema. Please file an issue with the owner of the MCP server.
  

• v1.0.23 (August 14, 2025)

  1. A tool get_attachment_files_from_Gmail was updated. Ref

• v1.0.24 (August 14, 2025)

  I’ve dealt with an issue caused by the update of the Gemini CLI from v0.1.18 to v0.1.19.

  The update has lowered the schema nesting depth, which now requires the use of $ref. However, the parametersJsonSchema appears to be incompatible with this. This means that a request body, such as the batchUpdate for the Docs, Sheets, and Slides APIs, cannot be directly generated using the input schema. It would be very useful if an AI agent could directly generate this type of complex request body.

  To solve this, I explored solutions to manage the low depth without using $ref for these three tools. I came up with the following workaround:

  The JSON schema is included in the tool's description, and this description is used as a prompt for the AI agent.
  

  With this workaround, the input schema remains simple, and the complex JSON schema can still be used to generate the request body. Additionally, no errors occur when using the Gemini CLI.

• v1.0.25 (August 14, 2025)

  1. Modified descriptions of manage_google_docs_using_docs_api, manage_google_sheets_using_sheets_api, and manage_google_slides_using_slides_api.

• v1.0.26 (August 18, 2025)

  1. Following 6 tools were added.
  • comments_drive_api_list: Use to get a list of a file's comments on Google Drive.
  • comments_drive_api_remove: Use to delete a comment using the "comments.delete" method of Google Drive API.
  • revisions_drive_api_list: Use to get a list of a file's revisions on Google Drive.
  • drive_activity_api_query: Use to query past activity in Google Drive. The activities of the files and folders in Google Drive are retrieved.
  • people_otherContacts_search: Use to provide a list of contacts in the authenticated user's other contacts that matches the search query.
  • people_people_getBatchGet: Use to provide information about a list of specific people by specifying a list of requested resource names.

• v1.0.27 (August 22, 2025)

  1. Following 6 tools for Google Analytics were added.
  • analytics_admin_accountSummaries_list: Retrieves a list of all Google Analytics accounts accessible by the current user
  • analytics_admin_properties_get: Get detailed information about a single Google Analytics property
  • analytics_data_properties_runReport: Fetches a custom report from a Google Analytics property
  • analytics_data_properties_runRealtimeReport: Generates a customized report of real-time event data from a Google Analytics property

TOP