Why I Built a DevSecOps Portfolio with a Fake Hacker Terminal and Live Attack Logs
Most portfolio websites feel like glorified PDF resumes. You get the same sterile minimalist layout, three cards showing Todo apps, and an "About Me" section that puts you to sleep. As a cybersecurity engineering studen
Most portfolio websites feel like glorified PDF resumes. You get the same sterile minimalist layout, three cards showing Todo apps, and an "About Me" section that puts you to sleep.
As a cybersecurity engineering student, I wanted something that actually reflects what I do all day: breaking stuff, looking at terminal logs, and trying to write secure code.
So I spent the last few weeks building my own interactive portfolio from scratch. Here is a breakdown of what I built, how it works, and why I intentionally made it log security events in real-time.
The Concept: Two Personalities (Terminal vs. Dashboard)
I couldnβt decide whether I wanted my portfolio to look like a gritty CLI hacker terminal or a high-contrast Neo-Brutalist dashboard.
So I decided: why not both?
I built a viewport controller with a physical switch:
-
Terminal Mode: An interactive mock Linux shell. You can run commands like
help,cat about.txt,skills,projects, andwhoami. It has autocomplete, command history with the Up/Down arrow keys, and retro keyboard click sounds generated via the browser's native Web Audio API (no external MP3 bloat). -
Dashboard Mode: A bold Neo-Brutalist layout with hard 3px black borders, harsh drop shadows, high-contrast badges, and interactive cards for people who prefer a standard GUI.
The switch between modes preserves state and doesn't trigger any page reload.
---
## Feature Highlight: The Attack-Logging Guestbook
Most guestbooks on personal sites are just a simple
INSERT INTOform. Since this is a DevSecOps project, I turned the guestbook into a honeypot-lite experiment: - Input Sanitization & XSS Defense: Every submission is strictly validated using Zod schemas on the Next.js API layer. All incoming strings are stripped of HTML tags, script tags, and payload patterns.
-
Real-time Security Telemetry: When someone attempts an injection payload or triggers rate limits, the backend flags the attempt with a risk score, logs the event to a PostgreSQL database (hosted on Neon Serverless), and pushes the event to a public
Security Logsdashboard.
3. No External Sound Libraries: I wanted auditory feedback for button presses and terminal keystrokes. Instead of importing heavy audio assets, I used the Web Audio API to synthesise sine and square waves directly on the client.
The Stack
- Framework: Next.js (App Router, TypeScript)
- Styling: Tailwind CSS with custom Neo-Brutalist tokens
- Database & ORM: PostgreSQL + Prisma ORM
- Hosting & Infrastructure: Vercel + Cloudflare DNS
- Sound Engine: Native browser AudioContext
Check It Out
The portfolio is live right now. Feel free to try out the terminal or test the guestbook inputs:
- π Live Website: clean-code.my.id
- π GitHub: Christian Luis Paskalis Ginting (christianLuis07) Iβd love to hear your feedbackβespecially if you can find creative ways to bypass the sanitization layer! --- Written by Christian Luis Paskalis Ginting β Cybersecurity Engineering student & full-stack developer.
Originally published by Dev.to WebDev. Aggregated on AIWithGhost for educational purposes β full credit and traffic to the original publisher.