The Attacker's Mindset (Part 3/3): Running a 45-Minute Cyber Attack Challenge
Part 3 of 3: the final session of the offensive security training I gave at the Algérie Télécom Skills Center in Saïda (7-77 Digital Skills Programme). Part 1 was the mindset, part 2 the hands-on lab. This is where every
Part 3 of 3: the final session of the offensive security training I gave at the Algérie Télécom Skills Center in Saïda (7-77 Digital Skills Programme). Part 1 was the mindset, part 2 the hands-on lab. This is where everyone had to prove it.
No lectures. Just targets, a scoreboard and a clock.
The rules
Teams of 2 or 3 (no laptop? pair up), and three phases, each on a timer.
1. Setup: 10 minutes. Get Juice Shop running and open it through Burp. A working local lab earned the first points.
docker run --rm -p 3000:3000 bkimminich/juice-shop
2. Hunt: 45 minutes. No step-by-step. Recon, question every input, exploit, in any order, at your own pace. Found something? Raise a hand, show it, and explain how you got there.
3. Report: 10 minutes. Pick your best findings and write a short report for each.
First move: recon
Juice Shop hides a Score Board that tracks every challenge and its points. It isn't in the menu. Finding it is the first real task, and it's recon in its purest form: discovering what the site doesn't advertise. Once a team found it, the whole hunt opened up.
What teams hunted
- Normal XSS: make the site run your code (a popup).
- XSS with real impact: go past the popup and expose the victim's session token.
- SQL injection: log in as admin with no password.
- Crack the admin's password: decode the token, follow the trail, crack it.
- Read the code: on a challenge you solved, find the vulnerable line and write the fix.
How scoring worked
Every finding was scored on three things: exploit + explain + report, never on the payload alone, with a bonus for clean methodology.
That was deliberate. In a real job nobody pays for a payload that works by accident. They pay for someone who can explain why it works and how to fix it.
The deliverable: the report
Each report had four parts:
- What: the vulnerability, in one sentence.
- Steps: exactly how to reproduce it.
- Impact: what an attacker gains.
- Fix: what the developer should change.
AI was allowed for drafting, with one rule: check every line before you submit it.
One week, zero to this
That was the arc of the week. Day 1: understand how attackers think. Day 2: find bugs by hand. Day 3: find, prove and report them as a team, against the clock.
Everything from all three days stays online and free for anyone who wants to keep going:
- 🔗 Workshop 3 challenge deck: https://medjahdi.dev/src/workshop-3
- 🎁 All slides, setup guides, payloads and the roadmap: https://medjahdi.dev/src
- ⭐ GitHub repository: https://github.com/medjahdi/skills-center-gift
Thanks to the team at the Algérie Télécom Skills Center in Saïda for hosting it.
Keep the mindset. Follow the roadmap. This was only the beginning.
Originally published by Dev.to Security. Aggregated on AIWithGhost for educational purposes — full credit and traffic to the original publisher.