competition-android-hooking
Pass
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were identified in the skill instructions or configuration files. All provided guidance aligns with standard security auditing practices.\n- [PROMPT_INJECTION]: No attempts to override system prompts or bypass safety guidelines were detected. The localization instruction to use Simplified Chinese is a standard functional requirement.\n- [DATA_EXFILTRATION]: No evidence of hardcoded credentials or unauthorized network operations was found. Sensitive file paths mentioned in the text (such as databases and shared preferences) refer to the subject of the analysis (the Android app) rather than the agent's environment.\n- [REMOTE_CODE_EXECUTION]: No patterns indicating the download or execution of remote code from untrusted sources were found. The skill relies on local reference files and standard analysis workflows.
Audit Metadata