mobile-design

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION] (LOW): The skill utilizes 'MANDATORY Checkpoints' and 'Yasak Listesi' (Forbidden List) to steer AI behavior. While these are prompt steering techniques, they are used to enforce development best practices and do not target safety filter bypass or role-play injection.
  • [DATA_EXFILTRATION] (SAFE): No exfiltration patterns detected. The skill specifically includes security guidelines to prevent data exposure, such as advising against storing tokens in AsyncStorage and suggesting SecureStore or Keychain instead.
  • [EXTERNAL_DOWNLOADS] (SAFE): The skill references standard development tools (Flipper, Reactotron, Proxyman) and frameworks (Zustand, Redux). No malicious external script downloads or piped execution patterns (e.g., curl|bash) were found.
  • [COMMAND_EXECUTION] (SAFE): Mentions standard mobile development and debugging commands such as adb logcat, pod install, and ./gradlew clean. These are legitimate and expected within the context of the skill.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The skill uses a 'CHECKPOINT' mechanism where it expects the AI to reflect on its work. While this is an ingestion point for untrusted user input, the skill itself provides the boundary markers and instructions to maintain quality, representing a low risk surface.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:48 PM