Kordant/tasks/ios-production/06-jailbreak-detection.md

06. Jailbreak Detection & Runtime Security

meta: id: ios-production-06 feature: ios-production priority: P2 depends_on: [] tags: [security, hardening, production]

objective:

• Implement jailbreak detection and runtime security measures to protect the app on compromised devices

deliverables:

• Jailbreak detection implementation
• Runtime integrity checks
• Anti-tampering measures
• Secure enclave usage for sensitive operations

steps:

1. Implement jailbreak detection:
   • Check for common jailbreak files (/Applications/Cydia.app, etc.)
   • Check if app can write outside sandbox
   • Check for suspicious dylibs
   • Use multiple detection methods for robustness
   • Add to APIClient or AppDelegate
2. Define jailbreak response:
   • Option A: Block app usage with warning
   • Option B: Degrade functionality (no biometric, no payments)
   • Option C: Log and alert backend
   • Recommended: Option B + alert backend
3. Implement runtime integrity checks:
   • Verify code signature at runtime
   • Detect debugger attachment
   • Detect code injection attempts
   • Verify method swizzling hasn't occurred
4. Use Secure Enclave:
   • Store encryption keys in Secure Enclave
   • Use biometrics via LocalAuthentication framework
   • Protect keychain items with biometry constraint
5. Add anti-tampering:
   • Obfuscate sensitive strings (API endpoints, keys)
   • Verify bundle identifier hasn't changed
   • Check for binary modification
6. Implement backend alerting:
   • Send jailbreak detection event to backend
   • Include device info (non-identifiable)
   • Flag account for additional monitoring

tests:

• Unit: Test detection logic with mock jailbreak indicators
• Integration: Test on jailbroken device (if available)
• Security: Verify debugger detection works

acceptance_criteria:

• Jailbreak detection active with multiple methods
• App degrades gracefully on detected jailbreak
• Backend receives alert when jailbreak detected
• Secure Enclave used for key storage
• Debugger attachment detected and handled
• Runtime integrity checks active
• Sensitive strings obfuscated in binary
• No false positives on non-jailbroken devices

validation:

• Run on normal device → no jailbreak detected, full functionality
• Run on jailbroken device → degraded mode activated
• Attach debugger → app detects and responds
• Check backend logs → jailbreak events received

notes:

• Jailbreak detection is cat-and-mouse — don't rely on it exclusively
• Apple may reject apps that overly aggressively block jailbroken devices
• Degradation is safer than blocking (better user experience)
• Use Swift string obfuscation libraries for sensitive data