Google Drive Server

Search and access files in Google Drive

Overview

The Google Drive server enables AI agents to search and access files stored in Google Drive. It provides read-only access to files and supports various file types.

Prerequisites

• Python 3.11+
• A Google Cloud Project with Google Drive API enabled
• OAuth 2.0 credentials with scope: https://www.googleapis.com/auth/drive.readonly

Authentication

Setup OAuth Credentials

1. Create a new Google Cloud project
2. Enable the Google Drive API
3. Configure an OAuth consent screen
4. Add OAuth scope: https://www.googleapis.com/auth/drive.readonly
5. Create an OAuth Client ID for “Desktop App”
6. Save credentials as local_auth/oauth_configs/gdrive/oauth.json

Authenticate

python src/servers/gdrive/main.py auth

Available Tools

search

Description: Search for files in Google Drive

Parameters:

• query (string, required): Search query to find files

Example:

{
  "query": "project proposal"
}

Returns: List of matching files with names, IDs, MIME types, and metadata

Resources

The Google Drive server provides access to file resources:

List Resources

Lists files from your Google Drive (up to 10 at a time).

URI Format: gdrive:///{file_id}

Read Resource

Reads file content with automatic format handling:

Google Workspace Files

Automatically exports Google Workspace files to readable formats:

File Type	Export Format
Google Docs	Markdown
Google Sheets	CSV
Google Slides	Plain text
Google Drawings	PNG

Regular Files

• Text files (.txt, .md, .json) are returned as UTF-8 text
• Binary files are returned as-is

Usage Examples

Searching for Files

# Search for all PDFs
files = await call_tool("search", {
    "query": "type:pdf"
})

# Search by name
files = await call_tool("search", {
    "query": "annual report 2024"
})

# Search within a specific folder (requires folder ID)
files = await call_tool("search", {
    "query": "'folder_id' in parents"
})

Reading Files

# List available files
files = await list_resources()

# Read a specific file
for file in files:
    content = await read_resource(file.uri)
    print(f"File: {file.name}")
    print(f"Content: {content}")

Accessing Google Docs

# Search for Google Docs
docs = await call_tool("search", {
    "query": "type:document"
})

# Read the document (automatically exports to markdown)
for doc in docs:
    content = await read_resource(f"gdrive:///{doc['id']}")
    # Content is now in markdown format

Google Drive Search Syntax

The search tool supports Google Drive’s query syntax:

Query	Description	Example
type:	File type	type:pdf, type:document
title:	File name	title:"Project Plan"
fullText:	Content search	fullText:"quarterly results"
modifiedDate	Date modified	modifiedDate > '2024-01-01'
in parents	In folder	'folder_id' in parents
trashed = false	Not in trash	trashed = false

Common File Types

• type:document - Google Docs
• type:spreadsheet - Google Sheets
• type:presentation - Google Slides
• type:pdf - PDF files
• type:folder - Folders

Running the Server

Local Development

python src/servers/local.py --server gdrive --user-id local

Best Practices

1. Specific queries: Use specific search terms to reduce results
2. File types: Filter by file type to narrow searches
3. Read-only: This server provides read-only access
4. Export formats: Google Workspace files are automatically converted

API Reference

Tool	Purpose	Common Use Cases
search	Find files	Locate documents, find specific file types

Supported MIME Types

Google Workspace files are automatically exported:

• application/vnd.google-apps.document → text/markdown
• application/vnd.google-apps.spreadsheet → text/csv
• application/vnd.google-apps.presentation → text/plain
• application/vnd.google-apps.drawing → image/png

Limitations

• Read-only access (no file uploads or modifications)
• Binary files larger than available memory may cause issues
• Some file formats may not export perfectly

Related Resources

• Google Drive API Documentation
• Drive Search Query Terms