EmailExtractor

Automated pipeline that fetches technical article digests from emails, extracts URLs, and performs AI-powered analysis using Gemini CLI.

Features

• Email Fetching: Connect to IMAP servers and filter unread emails by sender/subject
• URL Extraction: Extract and clean article URLs from email bodies (optimized for Medium digests)
• AI Analysis: Deep technical analysis of articles via Gemini CLI
• Incremental Reporting: Generate Markdown and JSON reports with auto-save

Quick Start

# Create conda environment
conda env create -f environment.yml
conda activate emailextractor

# Configure credentials
cp .env.example .env
# Edit .env with your email app password

# Edit config.toml with your IMAP settings and filters

# Run
python main.py

Configuration

Edit config.toml to configure:

• IMAP settings: Server, port, username
• Filters: Email sender, subject keywords, max emails per source
• Output: Directory, prompt file, analysis timeout

Use .env for sensitive credentials (EMAIL_PASSWORD).

CLI Options

python main.py --help

python main.py --fetch-url --analyze-url    # Full pipeline (default)
python main.py --no-analyze-url             # Fetch only
python main.py --analyze-url --input-file report.json  # Analyze existing report

Project Structure

src/
├── config.py    # Configuration management
├── fetcher.py   # IMAP email fetching
├── parser.py    # URL extraction
├── analyzer.py  # Gemini CLI integration
└── reporter.py  # Report generation

Requirements

• Python 3.10+
• Gemini CLI installed and configured
• IMAP access to email account