node-chatgpt-email-spam-detect

This sample will show you to easily use the Nylas Node.js SDK and the OpenAI Node.js SDK to categorize your email messages.

Setup

System dependencies

• Node.js v16.x

Gather environment variables

You'll need the following values (visit the Nylas dashboard at dashboard.nylas.com):

CLIENT_ID = "" # You can get this from the Nylas dashboard
CLIENT_SECRET = "" #  You can get this from the Nylas dashboard
ACCESS_TOKEN = "" # You can get this by authing a user account (e.g. via the Nylas dashboard)
RECIPIENT_ADDRESS = "" # The auth user account's email address
OPENAI_API_KEY = "" # You can get this from the OpenAI dashboard

Add the above values to a new .env file:

$ touch .env # Then add your env variables

Install dependencies

$ npm i

Usage

Run the script using the node command:

$ node index.js

When your messages are successfully read, you'll get the following output in your terminal for a few emails:

# date, subject, category, and message ID
[4/8/YYYY] Here's an inportant message - Yes, security. (1yz01ivndb)
[4/7/YYYY] Here's an email subject - No, newsletters. (2wwqyz01ivnmzb)
[4/6/YYYY] Another subject - No, spam. ((241ahbvvivnmzb))

The core of this example is the "category" in the output above. This is the category that the OpenAI API has assigned to the email message, formatted as "Yes, {category}" or "No, {category}". The "Yes" or "No" indicates whether the message is likely to be important to you.

Learn more

Visit our Nylas Node.js SDK documentation to learn more.

Expand on this example

• Give the prompt the option to be "unsure"
• Let the user "train" the prompt by storing an allow/block list to add to the prompt
• Expose the results in a graphical email client
• Use the results to categorize emails in Gmail/Outlook/Exchange
• Expand on the list of categories to triage