Personal Portfolio Website

A fully static, dark-themed personal portfolio website hosted on GitHub Pages. It dynamically fetches and displays:

• GitHub Repositories: Public repos sorted by recent activity via GitHub API.
• Google Scholar Publications: Pulled monthly via a Node.js CLI into data/gscholar.json.
• LinkedIn Profile Data: Fetched manually via a Python CLI into data/linkedin.json, including education, experiences, and top skills.

Repository Structure

/ (root)
├── index.html           # Main page
├── css/                 # Stylesheets
│   └── styles.css
├── js/                  # Front-end scripts
│   └── scripts.js
├── data/                # Fetched JSON data
│   ├── gscholar.json
│   └── linkedin.json
├── scripts/             # Data fetcher CLIs
│   ├── fetch_scholar.js
│   └── fetch_linkedin.py
├── .github/             # GitHub Actions workflows
│   ├── workflows/
│   │   └── update-gscholar-monthly.yml
├── package.json         # Node.js project config
├── .gitignore
└── README.md            # This file

Getting Started

1. Install Dependencies

   • Node.js: npm install (for the Scholar scraper)
   • Python 3: Set up a virtual environment (Python >= 3.10 is required by StaffSpy) and install dependencies:

     python3.11 -m venv .venv
     source .venv/bin/activate
     pip install "staffspy[browser]"

2. Fetch Data Locally

   npm run fetch-scholar               # Update Google Scholar publications
   python scripts/fetch_linkedin.py    # Update LinkedIn profile data (will open a browser on first run)

3. Serve or Build

   • The site is static—simply open index.html or deploy to GitHub Pages.

Automating LinkedIn Refresh on GitHub Actions

Since LinkedIn requires browser authentication, GitHub Actions runs the scraper using your local session cookies stored in data/session.pkl. To set this up:

1. Generate the Session File Locally Run the LinkedIn script locally:

   python scripts/fetch_linkedin.py

   A browser window will open. Log into your LinkedIn account. Once logged in, go back to your terminal and press Enter to let the script finish and save data/session.pkl locally.

2. Base64 Encode the Session File Run the following command to encode data/session.pkl into a base64 string and copy it:

   • macOS:

     base64 -i data/session.pkl | pbcopy

   • Linux / Git Bash:

     base64 -w 0 data/session.pkl

3. Add GitHub Secret Go to your GitHub repository: Settings > Secrets and variables > Actions > New repository secret.

   • Name: LINKEDIN_SESSION_B64
   • Secret: Paste the base64 string you copied.

The monthly workflow will automatically decode this session cookie to run the scraper in non-interactive headless mode. If the session expires, the workflow will continue safely without blocking Google Scholar updates, and you will see an EOFError in the action log indicating you need to regenerate the session secret.

Site Sections

• Hero: Profile photo, name, and headline.
• About: Biography pulled from LinkedIn.
• Education: Academic history list.
• Work: Experience cards, location, dates.
• Skills: Top 3 skills as badges.
• Repositories: Latest GitHub repos.
• Publications: Key papers with metadata.
• Contact: Email (from LinkedIn), GitHub, LinkedIn, Scholar, ORCID.

---

Built with HTML, CSS, JavaScript (vanilla), Node.js, Python, and GitHub Actions.