📧 EmailAI

An AI-powered email agent that syncs your emails via IMAP and provides intelligent analysis using OpenAI, Anthropic, or local Ollama.

✨ Features

• 🔄 IMAP Sync - Mirror your email account locally with batch syncing
• 👁️ IDLE Support - Real-time email detection (when server supports it)
• 🤖 AI Analysis - Summarize, categorize, and extract insights from emails
• 📤 SMTP Send - Send replies generated by AI
• 🔍 Search - Full-text search across all synced emails
• 📦 Local Storage - Emails stored as JSON, easy to inspect and backup
• 🔧 Configurable - Works with Gmail, Outlook, or any IMAP server

🚀 Quick Start

Prerequisites

• Go 1.21 or higher
• An email account with IMAP access enabled
• (Optional) AI API key for smart features

Installation

# Clone the repository
git clone https://github.com/peterretief/emailai.git
cd emailai

# Build
go build -o emailai .

# Or install directly
go install github.com/peterretief/emailai@latest

Configuration

# Copy the example config
cp config.example.json config.json

# Edit with your settings
nano config.json

Gmail Setup

1. Enable 2-Step Verification at https://myaccount.google.com/security
2. Generate an App Password at https://myaccount.google.com/apppasswords
3. Use the 16-character app password (not your regular password)

{
  "imap": {
    "host": "imap.gmail.com",
    "port": 993,
    "username": "you@gmail.com",
    "password": "your-app-password",
    "tls": true
  },
  "smtp": {
    "host": "smtp.gmail.com",
    "port": 587,
    "username": "you@gmail.com",
    "password": "your-app-password",
    "tls": true
  }
}

AI Setup (Optional)

Choose one of the supported AI providers:

OpenAI:

{
  "ai": {
    "enabled": true,
    "provider": "openai",
    "api_key": "sk-your-key-here",
    "model": "gpt-4"
  }
}

Anthropic:

{
  "ai": {
    "enabled": true,
    "provider": "anthropic",
    "api_key": "sk-ant-your-key-here",
    "model": "claude-3-opus-20240229"
  }
}

Ollama (Local, free):

Install Ollama from https://ollama.com, then pull a local model:

ollama pull llama3.1:8b

Make sure the Ollama server is running:

ollama serve

Configure EmailAI to use the local model:

{
  "ai": {
    "enabled": true,
    "provider": "ollama",
    "api_key": "",
    "model": "llama3.1:8b",
    "max_tokens": 2048
  }
}

Good free local model choices include llama3.1:8b, qwen2.5:7b, and mistral:7b. For email analysis and structured extraction, qwen2.5:7b is also worth trying:

ollama pull qwen2.5:7b

Then set ai.model to qwen2.5:7b.

Usage

# Sync all emails
./emailai -action sync

# Sync specific folder
./emailai -action sync -folder INBOX

# Force a complete rescan when needed
./emailai -action sync -full-sync

# Run as continuous sync server
./emailai -action serve

# Analyze emails with AI
./emailai -action analyze -query "meeting"

# List contacts found in downloaded mail
./emailai -action contacts -limit 100

# View sync stats
./emailai -action stats

# Search email text and locally extracted OCR text
./emailai -action search -folder INBOX -query "driver licence"

# Restrict the search to image attachments
./emailai -action search -folder INBOX -attach jpg -query "driver licence"

# Send test email
./emailai -action send

For large Gmail mailboxes, increase the batch size in config.json so each sync run imports more messages:

{
  "sync": {
    "maildir_path": "",
    "store_path": "./emails",
    "max_batch_size": 1000
  }
}

Run sync repeatedly until each folder is up to date:

while ./emailai -action sync; do sleep 10; done

Press Ctrl+C to stop the loop.

📁 Project Structure

emailai/
├── main.go              # Entry point and CLI
├── config/
│   └── config.go        # Configuration management
├── sync/
│   └── sync.go          # IMAP sync engine
├── store/
│   └── store.go         # Local email storage
├── ai/
│   └── ai.go            # AI processing (OpenAI/Anthropic/Ollama)
├── smtp/
│   └── smtp.go          # Email sending
├── config.example.json  # Example configuration
└── README.md

🔧 Configuration Options

When ocr.enabled is true, image attachments are processed locally with Tesseract during sync. Extracted text is stored in the local email JSON and included in search. Attachment image bytes are not stored by the OCR feature.

Normal syncs store a per-folder IMAP UID watermark and fetch only newer messages. Use -full-sync to rescan every message, for example after enabling OCR or changing message parsing.

Set filters.exclude_newsletters to true to keep obvious newsletters, promotions, and automated bulk mail out of search and analyze results. Tune the exclude_from, exclude_subject, and exclude_body lists for your mailbox.

To use mbsync/isync for mailbox synchronization, set sync.maildir_path to the Maildir root. When this is configured, the application reads local Maildir files and does not connect to IMAP. Run mbsync first, then run ./emailai -action sync to import new messages and process OCR.

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

⚠️ Security Notes

• Never commit your config.json file (it's in .gitignore)
• Use App Passwords, not your regular email password
• Keep your AI API keys secure
• The emails/ directory contains your synced emails - treat it as sensitive data

🙏 Acknowledgments

• go-imap - IMAP library for Go
• go-imap-idle - IDLE extension
• go-message - Email message parsing

📧 Support

If you have any questions or issues, please open an issue on GitHub.