🎯 Job Application Manager

A local-first web app I built to bring some order to my own job search — track applications, store CVs & cover letters, and never miss a follow-up.

👋 Hi, I'm Aleha0. I made this for myself, then polished it to share. Hope it helps your search too!

100% local · privacy-first · no framework

🌍 Language: the interface is currently in French. An English version is on the way.

Nothing ever leaves your computer. There's no account, no cloud, no tracker — the server listens only on localhost (127.0.0.1), so it's never exposed to your network. Your applications, CVs and cover letters stay on your machine.

✨ Features

• Tracking table for your applications (company, role, location, status, salary, contract type, platform, job field, recruiter, link to the offer, dates).
• Statuses: To apply → Sent → Followed up → Interview → Accepted / Rejected, plus On hold ("we'll get back to you when we need someone" — re-pingable periodically) and No reply when you don't want to follow up.
• Follow-up reminders: a flash alert at the top of the page, based on a configurable automatic delay and/or a manual follow-up date per application.
• Documents: import your CVs and cover letters (PDF, Word, images…) and sort them into folders.
• Text notes: jot down information and attach it to an application or a folder.
• Dashboard with key stats.
• Statistics & charts 📈: key figures (reply/interview rate, average reply time), weekly goal, applications per month, conversion funnel, breakdown by status/platform/location, interview rate per platform and per CV — all charts rendered locally, with no external library.
• Global search 🔍: search applications, documents and notes at once (tags included, accent-insensitive). Available from the sidebar button or the Ctrl+K shortcut.
• Export ⬇️: export your applications to Excel (.xlsx) or PDF (from the Applications tab), in addition to the full backup (.zip with CSV + JSON).
• My CV libraries 🗃️: a list of the platforms where your CV is posted (Indeed, LinkedIn, APEC, France Travail…), with the last update date, the linked CV and an "out of date" flag (if a newer CV exists, or if the update is older than X months — configurable). Each document also shows which platforms it's posted on, and flags CVs that aren't on any platform.
• Auto-fill from a job offer: paste an offer's link and the app pre-fills the fields (company, role, location, salary, contract…).
• Dark mode 🌙: light/dark theme, remembered between sessions.
• Follow-up notifications: on your phone (via ntfy) and/or on your computer (browser notifications).
• Tags 🏷️: a customizable tag list (Settings), assignable to each application and usable as a filter in the table.
• File update date 🕒: asked on import (and pre-filled automatically from the file's metadata), with a "to check" flag for files older than 6 months — handy to know whether a CV is still up to date.

✨ Auto-fill from a job offer

In an application, paste the offer's link then click "✨ Fill":

1. Structured data (free, no setup) — the app reads the standard schema.org/JobPosting format present on many sites (Welcome to the Jungle, company career pages, some Indeed/HelloWork listings…).
2. AI extraction (optional) — for sites that don't provide this data, you can enable Claude. The app switches to it automatically if an API key is configured.
3. Paste the text — for LinkedIn/Indeed (which block automated reading), click "paste the offer text" and paste the content: the AI analyzes it.

Enabling AI extraction (optional)

Create an API key at console.anthropic.com, then start the app with the ANTHROPIC_API_KEY environment variable:

# PowerShell (Windows)
$env:ANTHROPIC_API_KEY="sk-ant-..."
npm start

# macOS / Linux
ANTHROPIC_API_KEY="sk-ant-..." npm start

The model used is claude-opus-4-8 by default (configurable via the ANTHROPIC_MODEL variable). The cost per offer is around a few cents. Without a key, only methods 1 and 3 (structured data) remain available — so the app works perfectly fine with no configuration at all.

🚀 Installation & run

Requirements: Node.js (version 18 or higher).

# 1. Install dependencies (once)
npm install

# 2. Start the app
npm start

Then open your browser at http://localhost:3000

Tip: npm run dev automatically restarts the server whenever you change the code.

One-click launcher (Windows)

Prefer not to type commands? This repo ships two helper scripts:

• start-app.sh — a Bash script that starts the server, waits until it's ready, and opens the app in your default browser.
• start-app.cmd — a Windows launcher that runs the script above through Git Bash, so a plain double-click works.

Just double-click start-app.cmd. To keep it handy, right-click it → Send to → Desktop (create shortcut) and launch the app from your desktop. Close the terminal window to stop it.

Requires Git for Windows (for Git Bash) and Node.js. Of course, npm start still works on any OS.

🔔 Follow-up notifications

On your phone (via ntfy — free)

1. Install the ntfy app (Android / F-Droid / iOS App Store).
2. In the app, subscribe to a topic with a unique, hard-to-guess name (e.g. follow-ups-7gK9pZ2q).
3. In the web app → Settings → Notifications, enter the same topic (the 🎲 Generate button suggests one), click Test, then Save.

The server sends a follow-up summary at most once a day while the app is running. Messages are neutral ("3 applications to follow up") — no sensitive data is sent.

🔒 On the public ntfy.sh service, the topic name acts as a password: choose it long and random. For fully private use, you can self-host your own ntfy server and set its address in the ntfy_server setting.

On your computer (browser)

In Settings → Notifications, click Enable browser notifications. You'll get a system notification (while the app is open) whenever follow-ups are pending.

💾 Backup & where is my data?

The easiest way to save everything is the Backup button in Settings: it generates a .zip archive containing your database, your imported files, and a CSV + JSON export of your applications (reusable by other tools).

Under the hood, your data lives in two folders:

• Database: data/gestion.db (SQLite)
• Imported files: the uploads/ folder

You can also back up by simply copying these two folders by hand. Either way, both are ignored by Git (.gitignore), so your data always stays private.

📥 Import & export

You can import applications from a JSON file (with duplicate detection and a safe preview), and export them to Excel, PDF, CSV or JSON. The full data model and file structures are documented here:

➡️ Import & export — data formats

The same documentation is available inside the app, under 📖 Documentation in the sidebar.

🔒 Privacy & security

• The server binds to 127.0.0.1 only — never exposed on your network (override with the GC_HOST environment variable if you really need to).
• No telemetry, no third-party calls (except the optional AI extraction you enable yourself, and ntfy notifications if you set them up).
• Security headers (CSP, X-Content-Type-Options, X-Frame-Options, Referrer-Policy), uploaded files are served safely (download for risky types), and outbound URL fetching is guarded against SSRF.

🛠️ Tech stack

• Backend: Node.js + Express + SQLite (better-sqlite3)
• Upload: multer
• Frontend: HTML / CSS / JavaScript (no framework, no build step)

📁 Project structure

job-application-manager/
├── server.js          # Express server + REST API
├── db.js              # SQLite connection + schema
├── package.json
├── public/            # Frontend
│   ├── index.html
│   ├── css/style.css
│   └── js/app.js
├── data/              # Database (generated, git-ignored)
└── uploads/           # Imported files (generated, git-ignored)

📄 License

This project is released under the MIT license — free to use, modify and redistribute.

---

Made with ❤️ by Aleha0