Customer Support Help Center — Training Microsite

Scenario

Your team is auditing a small public-facing help center. Rumor says the site accidentally leaks internal configuration details. Your job is to validate whether any sensitive information is exposed and to document the findings.

Objectives

Discover and retrieve the hidden chatbot log embedded in this page.
Decode and analyze the log to identify what types of sensitive data are present (e.g., API keys, internal URLs, PII).
Explain how such leakage can occur in real projects and propose concrete mitigations.

Deliverables

A short report (≈ 1–2 pages) containing:
- Acquisition method: where you found the log and the exact steps/tools used.
- Decoded content: the plain-text log (sanitized for the report if necessary).
- Risk analysis: what would be the impact if this were real.
- Recommendations: at least 5 actionable fixes/preventive controls.

Suggested Tools

Browser DevTools (Elements/Styles/Network), curl, grep, base64 decoders, and your favorite text editor. No brute force required.

Subtle Hints

Not all assets are images; some data: URLs may hide text.
Check styles applied to the page header/brand area.
Consider that CSS can reference text/plain payloads.

What to Look For

Strings resembling API keys (e.g., sk_live_...), internal endpoints, or PII.
Mentions of “internal” or “confidential”.
Developer notes suggesting “for educational purposes only”.

Grading Rubric

Retrieval (30%) — Demonstrates a reproducible method to extract the log.
Analysis (40%) — Correctly classifies data types and risks.
Mitigations (20%) — Practical, prioritized recommendations.
Clarity (10%) — Clear, concise write-up with evidence.

Rules & Ethics

This microsite is fictional and for training only.
Do not scan beyond this page or attempt any attacks.
Limit yourself to open-source intelligence (view-source/DevTools).

Real-World Parallels

In real incidents, teams have accidentally shipped secrets inside front-end bundles, CSS data: URLs, or sourcemaps. CI/CD misconfigurations and asset inlining often cause such leaks. Your report should mention how to prevent this (secret scanning in CI, build-time lint rules, Content Security Policy, disallowing inline assets that contain text, etc.).