JASS - Job Application Support System

A local web application for managing job applications with AI-powered resume tailoring and cover letter generation.

Features

• Job Search: Search Greenhouse job boards by keywords and location
• Manual Job Entry: Paste job descriptions from LinkedIn or any source with auto-extraction of:
  • Job title, company, location
  • Salary range
  • Required skills
  • Remote/hybrid status
  • Hiring manager name (LinkedIn)
  • Job posting date
• AI Resume Tailoring: Generate tailored resumes using:
  • Claude API (Anthropic)
  • OpenAI API
  • Claude CLI (local, no API key needed)
• Cover Letter Generation: AI-generated cover letters with hiring manager personalization
• PDF Generation: Professional PDF output using md-to-pdf
• Application Tracking: Track job status (saved, ready, applied)
• Customizable Prompts: Edit AI prompts for resume/cover letter generation

Requirements

• Python 3.8+
• Node.js 18+ (for PDF generation)
• One of:
  • Claude API key (ANTHROPIC_API_KEY)
  • OpenAI API key (OPENAI_API_KEY)
  • Claude CLI installed

Linux/WSL Additional Requirements

PDF generation uses Puppeteer (headless Chrome). On Linux/WSL, you need Chrome dependencies:

sudo apt install -y libnspr4 libnss3 libatk1.0-0t64 libatk-bridge2.0-0t64 \
  libcups2t64 libdrm2 libxkbcommon0 libxcomposite1 libxdamage1 \
  libxfixes3 libxrandr2 libgbm1 libasound2t64

Note: Package names may vary by distribution. On older Ubuntu/Debian, use libatk1.0-0, libcups2, libasound2 instead of the t64 variants.

Quick Start

Windows

start.bat

Linux/macOS

chmod +x start.sh
./start.sh

The startup script will:

1. Create a Python virtual environment
2. Install Python dependencies
3. Install md-to-pdf (Node.js)
4. Start the Flask server at http://127.0.0.1:5000

Manual Installation

# Create virtual environment
python -m venv venv

# Activate (Windows)
venv\Scripts\activate

# Activate (Linux/macOS)
source venv/bin/activate

# Install Python dependencies
pip install -r requirements.txt

# Install md-to-pdf
npm install md-to-pdf

# Run
python app.py

Configuration

AI Provider Setup

1. Go to Settings in the web UI
2. Select your AI provider:
   • Claude API: Enter your Anthropic API key
   • OpenAI: Enter your OpenAI API key
   • Claude CLI: No API key needed (uses local Claude Code installation)
3. Select the model

Master Resume

1. Go to Resumes in the web UI
2. Upload or paste your master resume in Markdown format
3. Set as default for tailoring

Greenhouse Boards

1. Go to Settings
2. Add company board tokens (e.g., "cloudflare", "paloaltonetworks")
3. Use the search feature to find jobs

Usage

Adding Jobs

1. From Greenhouse: Use the Search feature to find and save jobs
2. Manual Entry: Click "Add Job" and paste the job description
   • LinkedIn format auto-detected
   • Fields extracted automatically

Generating Documents

1. Open a saved job
2. Click "Generate Tailored Resume & Cover Letter"
3. Wait for AI processing (shows progress)
4. Review and download PDFs

Application Workflow

1. Save jobs you're interested in
2. Generate tailored documents
3. Review/edit if needed
4. Download PDFs for submission
5. Mark as Applied when submitted

Debug Mode

Run with debug logging:

python app.py -d      # INFO level
python app.py -dd     # DEBUG level
python app.py -ddd    # DEBUG with verbose output

File Structure

Jass/
├── app.py              # Flask application
├── models.py           # Database models
├── ai_service.py       # AI provider abstraction
├── claude_cli.py       # Claude CLI integration
├── job_parser.py       # Job description parser
├── document_gen.py     # PDF generation
├── greenhouse.py       # Greenhouse API client
├── config.py           # Configuration
├── logger.py           # Logging setup
├── templates/          # Jinja2 templates
├── data/
│   └── applications/   # Generated documents
├── start.bat           # Windows startup
├── start.sh            # Linux startup
└── requirements.txt    # Python dependencies

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.