AI Email Assistant

This is an intelligent email client that uses the Google Gemini API to automatically categorize and label your emails. It simulates the functionality of a modern desktop email application, providing a familiar interface for managing your inbox while leveraging the power of AI for smart organization.

Key Features

• AI-Powered Labeling: Automatically suggests labels for incoming emails based on their content.
• User Feedback Loop: Learns from your corrections. When you accept, reject, or manually add a label, this action is logged as training data to improve future suggestions.
• Automated Rules Engine: Create custom rules to automatically label emails or move them to specific folders (e.g., "If sender contains 'billing', add 'Finance' label").
• Junk/Spam Filtering: Uses AI with a configurable confidence threshold to identify and label junk mail.
• Contact Management: Recognizes known contacts and allows for quickly adding new ones.
• Modern UI: A responsive, dark-mode compatible interface built with React and Tailwind CSS.
• Advanced Filtering & Sorting: Powerful tools to find the exact emails you're looking for.

How to Run

This application is a frontend demo and runs directly in your browser. To start it locally, you will need to use a simple web server. We provide scripts to make this easy.

Prerequisites

1. A Google Gemini API Key: The app's AI features will not work without it.
2. Python 3: The start-up scripts use Python to create a simple web server. Most systems have it pre-installed. If not, download it from python.org.

Step 1: Set Your API Key

The application reads your Gemini API key from an environment variable named API_KEY. You must set this variable in your terminal session before running the start script.

For Windows (in Command Prompt):

This command sets the key for your current command prompt session only.

set API_KEY="YOUR_GEMINI_API_KEY"

(Replace YOUR_GEMINI_API_KEY with your actual key.)

Important: You must run start.bat from the same command prompt window where you ran the set command. Double-clicking the file will not work unless you have set API_KEY as a permanent system-wide environment variable.

For macOS or Linux (in Terminal):

export API_KEY="YOUR_GEMINI_API_KEY"

(Replace YOUR_GEMINI_API_KEY with your actual key.)

Step 2: Run the Correct Start Script

We have provided scripts for different operating systems. Please use the one that matches your system.

On Windows: From the same command prompt you used in Step 1, run the script:

.\\start.bat

The script will check your setup, start the server, and automatically open the application in your browser.

On macOS / Linux: First, make the script executable (you only need to do this once):

chmod +x ./start.sh

Then, run the script:

./start.sh

Step 3: View the Application

Once the script is running, it will print a message like Serving HTTP on ... port 8000. You can now open your web browser and go to:

http://localhost:8000

Project Structure

The project is organized to separate concerns, making it easier for both humans and AI assistants to navigate, understand, and extend the codebase.

• / (root): Contains the main entry points (index.html, index.tsx) and project configuration.
• /components: Contains all UI-related React components. See the components/README.md for a detailed breakdown.
• /services: Contains modules for interacting with external APIs (like Gemini) and handling business logic separate from the UI. See the services/README.md for more details.
• /docs: Contains high-level design and architecture documentation for the project.

This structured approach, combined with the detailed READMEs in each directory, provides a clear and unambiguous guide to the project's architecture and conventions.