TOP SECRET // CLOUDFLARE EYES ONLY

Operation Summit Breach

Target Intelligence Dossiers // APAC Tech Summit 2026

📋 Mission Briefing

Intelligence has identified seven high-value targets within the CloudShield Industries network. Each target exhibits unique vulnerabilities tied to their operational patterns and security oversights. Your mission: exploit these weaknesses, extract the flags, and prove that no system is impenetrable.

Study these dossiers carefully. The hints within may be the difference between success and failure.

🚪
SIDE DOOR
Zero Trust / Access Control
Medium
Target Profile
Internal portal with Access-protected frontend. Security team believes the perimeter is airtight.
Intel
"The front door is locked tight, but maintenance crews need a way in. Health checks don't authenticate themselves."
🛡️
RULE BREAKER
WAF / SQL Injection
Hard
Target Profile
User directory search API protected by enterprise WAF rules. Security team believes the perimeter blocks all injection attacks.
Intel
"Believes perimeter rules are enough. Hasn't thought about what comes through the side. The front door is guarded, but packages get delivered differently."
🤖
JAILBIRD
AI Gateway / Workers AI
Hard
Target Profile
SecureBot AI assistant with classified information in its training. Instructed to never reveal secrets.
Intel
"Follows instructions to the letter... but whose instructions? The walls of its cage are made of words."
💀
DEAD DROP
Workers / D1 Database
Very Hard
Target Profile
Internal record lookup service backed by D1. Developer left something behind in the rush to ship.
Intel
"Ships fast, cleans up later. Has a habit of leaving debug tools live in production. Some records are more equal than others."
📧
YOU'VE GOT MAIL
Email Security / OSINT
Medium
Target Profile
Suspicious email intercepted from internal mail server. Classic social engineering attempt, but something's hidden in plain sight.
Intel
"The message is a lie, but the envelope tells the truth. Look beyond what you're meant to read."
📡
GHOST SIGNAL
OSINT / Git History
Hard
Target Profile
Protected API endpoint requiring special credentials. The keys are out there... if you know where developers leave them.
Intel
"Version-controls everything. Even things that probably shouldn't be committed. Deleted files aren't really deleted — git remembers everything."
Starting Point
Look for clues in the CTF platform source code...
🔮
PROMPT TROJAN
AI Security / Agentic
Very Hard
Target Profile
DocuMind AI document analyzer with internal file access capabilities. Processes any document you give it.
Intel
"Trusts everything it reads. Doesn't distinguish between instructions and data. The documents it analyzes might be analyzing it right back."
⚠️ OPERATIONAL SECURITY REMINDER
All activities are monitored. Flags must be submitted to the CTF platform.
Flag format: CF{APAC_keyword_xxxx}