mobile-security
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill functions as a purely analytical reference. It does not include scripts, binaries, or instructions for the agent to download or execute external code. All security logic is embedded within the prompt text as pattern-matching instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted mobile application source code (Android, iOS, React Native, and Flutter) provided by the user.
- Ingestion points: The agent is instructed to scan various project files, including
AndroidManifest.xml,Info.plist, and source code files (*.kt, *.swift, *.dart, etc.). - Boundary markers: No specific delimiters or boundary warnings are provided in the workflow to isolate untrusted code from the agent's instructions.
- Capability inventory: The skill explicitly states that 'no external tool required' and that it 'uses code analysis only'. It contains no instructions for network access, file system writes, or subprocess execution, which significantly mitigates the risk of an injection attack leading to system exploitation.
- Sanitization: There are no descriptions of input sanitization or filtering, but the lack of active capabilities makes this surface low-risk.
Audit Metadata