Job Hunter AU

Automated job discovery for Australia. Upload your resume, AI parses your skills, then GitHub Actions scrapes 5 job boards daily and emails you ranked results.

Zero hosting required - just a profile file and a GitHub Actions cron job.

Demo

Screenshots

Landing Page	Profile Editor

Scoring Weights	Export JSON

GitHub Automation	Email Digest

How It Works

flowchart LR
    A[Clone repo] --> B[Run wizard]
    B --> C[Connect GitHub]
    C --> D[All automatic]
    D --> E[Daily emails]

    style D fill:#065f46,stroke:#059669,color:#fff
    style E fill:#92400e,stroke:#d97706,color:#fff

Loading

You - clone, configure, connect. The app - forks, commits profile, sets secrets, enables cron. Every day - scrapes 5 boards, scores jobs, emails you.

Quick Start

Do not fork this repo first. Clone it, run the wizard, and the app will fork and configure everything for you.

1. Clone and run the wizard

git clone https://github.com/elvistranhere/job-hunter.git
cd job-hunter/web
cp .env.example .env        # Add your GEMINI_API_KEY (optional - you can skip AI)
npm install
npm run dev                  # Opens on localhost:3006

2. Configure your profile

1. Upload your resume PDF (or click "Continue without AI" to set up manually)
2. Customize skill tiers, scoring weights, locations, and roles
3. Adjust search settings (max hours, results per search, seniority filters)

3. Connect GitHub and deploy

On the automation page, click Connect GitHub and authorize. The app will automatically:

• Fork this repo to your account
• Commit your profile.json to the repo root
• Set Gmail secrets (GMAIL_USER, GMAIL_APP_PASSWORD, EMAIL_TO)
• Enable the daily cron workflow and trigger the first run

You'll receive an email with ranked job results within ~15 minutes. After that, the cron runs daily at 9pm UTC (7am AEST).

Gmail App Password: You'll need one for the email digest. Go to Google Account > App Passwords, generate one for "Mail", and enter it during setup. Requires 2-Step Verification.

Alternative: CLI or manual setup

CLI Only (no web app)

cp profile.example.json profile.json   # Edit with your skills, locations, roles
uv sync
uv run python scrape.py --profile profile.json
GMAIL_USER=you@gmail.com GMAIL_APP_PASSWORD=xxxx EMAIL_TO=you@gmail.com \
  uv run python email_digest.py --profile profile.json

Manual GitHub Actions (without OAuth)

1. Fork this repo - click Fork
2. Commit profile.json to the repo root:

   git add -f profile.json && git commit -m "Add my profile" && git push

3. Set repository secrets in Settings > Secrets and variables > Actions:

   Secret	Value
   GMAIL_USER	Your Gmail address
   GMAIL_APP_PASSWORD	Gmail App Password (16-character)
   EMAIL_TO	Comma-separated recipient emails

4. Enable Actions - click "I understand my workflows, go ahead and enable them"
5. Uncomment the cron in .github/workflows/daily-jobs.yml:

   schedule:
     - cron: '0 21 * * *'   # 9pm UTC = 7am AEST

6. Run workflow - Actions > Daily Jobs > Run workflow

Job Boards

Board	Method
Indeed	HTTP scraping
Seek	Browser automation (nodriver)
Prosple	GraphQL API
GradConnection	HTTP scraping
LinkedIn	HTTP scraping

Scoring

Jobs are scored on 8 weighted dimensions (all configurable in profile.json):

Category	Base Points	What It Measures
Skills Match	55	Core/strong/peripheral skill matches + adjacency
Location	20	Preferred cities + remote bonus
Title Match	18	Job title vs. your professional identity
Culture	15	Remote-first, equity, flex hours, learning budget
Company Tier	12	Big Tech, AU Notable, Top Tech companies
Sponsorship	12	Visa sponsorship signals
Job Quality	12	Salary transparency, detailed descriptions
Recency	10	Newer postings score higher

profile.json Reference

The web app generates this file; CLI flags are optional overrides.

Field	Type	Default	Description
skills	array	-	Skills with tiers: core, strong, peripheral
titles	string[]	-	Professional titles for title matching
keywords	string[]	-	Search terms for job boards
locations	string[]	-	Australian cities to search
roles	string[]	-	Role titles to search for
weights	object	all 1.0	Scoring multipliers (0 = off, 2 = double)
maxHours	number	24	Max hours since posted
resultsPerSearch	number	20	Results per search per site
excludeSeniority	string[]	["senior","lead",...]	Seniority levels to filter out
minScore	number	20	Minimum score for email digest

Project Structure

profile.example.json        Starter profile template
scrape.py                   Orchestration + scoring engine
scrapers_au.py              Job board scrapers (5 sources)
email_digest.py             HTML email rendering + SMTP
web/                        Next.js app (profile wizard)
.github/workflows/          CI + daily automation

Environment Variables

Web app (web/.env):

• GEMINI_API_KEY - Google AI key for resume parsing (optional - you can skip AI)
• NEXT_PUBLIC_GITHUB_CLIENT_ID - OAuth client ID for automated setup. A default is included in .env.example; the app handles the rest

GitHub Actions secrets:

• GMAIL_USER - Gmail address
• GMAIL_APP_PASSWORD - Gmail app password
• EMAIL_TO - Comma-separated recipient emails

Contributing

See CONTRIBUTING.md.

License

MIT