JobKernel

A high-throughput job application workbench — customize, track, and grind through as many applications as possible. Tailor resumes, generate cover letters, and keep an accurate record of your search without any step getting in your way.

Product direction and the "nothing blocks you by default" principle: documentation/product-direction.md.

---

🚨 For AI Agents & Developers

PLEASE USE DOCKER FOR DEVELOPMENT.

All dependencies are managed inside Docker containers. Do not run pip install or npm install locally unless specifically debugging a local environment issue.

./start-docker.sh

Container	Name
Backend (FastAPI)	jobapplicationautomator-backend-1
Frontend (Vite/React)	jobapplicationautomator-frontend-1

To run a command inside the backend container:

docker exec -it jobapplicationautomator-backend-1 pip install <package_name>

---

Features

• 📋 Application Tracking — Kanban, list & table views with drag-and-drop status updates. Move anything to any stage, any time — no step gates
• ⚙️ Optional Guided Mode — off by default; opt in per stage under Settings → Workflow if you want readiness checks enforced
• 📄 Resume Tailoring — AI rewrites your resume to match each job description, preserving formatting
• ✉️ Cover Letter Generation — AI-drafted cover letters with one click
• 🗂️ Archiving — Archive applications instead of deleting them; toggle archived view on the dashboard
• 🏢 Company Logos — Upload logos per application; displayed in all dashboard views
• 🔌 Browser Extension — Clip job listings directly from LinkedIn and other job boards
• 👤 Profile Management — Store contact info, resume data, documents, and preferences
• 📊 Analytics — Visual breakdown of your application pipeline

---

Tech Stack

Layer	Technology
Frontend	React, Vite, Tailwind CSS
Backend	Python, FastAPI, SQLAlchemy, SQLite
AI	Google Gemini (configurable)
Document Processing	python-docx, BeautifulSoup4
Extension	Chrome Manifest V3

---

Getting Started

Prerequisites

• Docker & Docker Compose
• A Google AI Studio API key (Gemini)

1. Clone the repository

git clone https://github.com/rdannenbring/job-kernel.git
cd job-kernel

2. Configure the backend

backend/config.json is not committed to the repo (it contains your API key). Create it from the provided template:

cp backend/config.example.json backend/config.json

Then open backend/config.json and fill in your values:

{
    "default_resume_path": "/mnt/resume/YourResume.docx",
    "ai_config": {
        "provider": "gemini",
        "model": "gemini-2.5-flash",
        "api_key": "YOUR_GEMINI_API_KEY_HERE"
    },
    ...
}

Where to get a Gemini API key: Visit aistudio.google.com, sign in, and create a new API key under Get API key.

⚠️ Never commit config.json to git. It is listed in .gitignore for this reason. If you accidentally expose a key, revoke it immediately at aistudio.google.com and generate a new one.

3. Start the application

./start-docker.sh

The app will be available at http://localhost:5173.

---

Project Structure

job-kernel/
├── backend/
│   ├── main.py                  # FastAPI app & all API endpoints
│   ├── config.json              # ⚠️ Local only — never committed
│   ├── config.example.json      # ✅ Safe template — copy to config.json
│   ├── requirements.txt
│   └── services/
│       ├── ai_service.py        # Gemini integration
│       ├── database_service.py  # SQLite ORM & migrations
│       ├── document_service.py  # Resume/cover letter DOCX generation
│       └── scraper_service.py   # Job URL scraping
├── frontend/
│   ├── src/
│   │   ├── pages/               # Dashboard, Profile, ApplicationDetail, etc.
│   │   ├── components/          # Sidebar, Layout, shared UI
│   │   └── App.jsx              # Root component & routing
│   └── index.html
├── extension/                   # Chrome browser extension
│   ├── manifest.json
│   ├── sidepanel.html/js/css
│   ├── content.js
│   └── background.js
├── resources/
│   └── images/logos/            # Brand assets
├── start-docker.sh
└── README.md

---

Privacy & Security

• All data is processed and stored locally on your machine — nothing is sent to external servers except AI API calls (Gemini)
• config.json and the SQLite database (*.db) are excluded from version control via .gitignore
• Resume files are stored in backend/uploads/ — also excluded from version control

---

License

MIT License — For personal use