Gmail to Google Sheets Automation

Yadnyesh More Full Stack Developer Intern 14th January 2026

1) Project Overview

This Python automation reads UNREAD emails from Gmail inbox and logs them into Google Sheets with columns: From , Subject , Date , Content. Successfully processed emails with zero duplicates!

2) Architecture Diagram (Hand Drawn)

3) Step-by-Step Setup Instructions

1. Google Cloud Console Setup

a. Created a New Project in Google Cloude and named as gmail-sheet-intern b. In Api services Enable Gmail API , Google Sheets Api c. OAuth consent screen in that i added test user email id to test the sheet d. Create Credentials e. Download the Credentials file as a Json

2. Local Development Setup

git clone https://github.com/yadnyeshmore/gmail-to-sheets.git

pip install -r requirements.txt

Put credentials/credentials.json (downloaded from step 1)

python src/main.py

4) Technical Implementation Details

OAuth 2.0 Authentication Flow

1st Run:

• Browser opens → Gmail login → Permissions → token.json created
• Uses InstalledAppFlow.run_local_server()

Future Runs:

• token.json auto-refreshed (no browser needed)
• Credentials valid for weeks/months

Duplicate Prevention Logic

processed_emails.json stores Gmail Message IDs: ["19bbc887a426b6d6", "19bbc7922434074d", ...90 total]

Before processing each email: if msg_id in processed_emails: print("Skipping duplicate") continue

After successful sheet append: processed_emails.add(msg_id) json.dump(list(processed_emails), file)

5) Challenges Faced & How I Solved Them

Challenge 1: OAuth "Access Blocked" Error

PROBLEM: "Gmail Sheets Intern not verified" blocked access

SOLUTION:

OAuth Consent Screen → Added yadnyeshmore@gmail.com as Test User

Browser → Advanced → "Go to Gmail Sheets Intern (unsafe)" → Allow

Screenshot captured for proof!

text

📊 Proof of Execution

Screenshots (All in /proof/ folder):

Video Demo (2:47 minutes)

demo_video.mp4

Limitations of Current Solution

1. Plain Text Only: HTML emails show raw HTML tags
2. Single Machine State: JSON file doesn't sync across servers
3. No API Retry: Network failures stop execution
4. 500 Char Content Limit: Truncated to fit Sheets cell limits
5. Local State File: Could grow large over years

Bonus Features Ready (Uncomment to enable)

# Subject filtering (line 75 in main.py)
if "Invoice" not in email_data['subject']:
    continue  # Process invoices only

# HTML to plain text (add html2text library)
# content = html2text.html2text(raw_html)

# Timestamp logging
# print(f"[{datetime.now()}] Processed {len(new_rows)} emails")

📈 Results & Performance

Total Emails Processed: 90 Duplicate Entries: 0 Success Rate: 100% Sheet Rows: 91 (including header)