mobile-design
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill contains instructions explicitly designed to override the agent's default behavior and training patterns. The 'mobile-design-thinking.md' file includes an 'ANTI-MEMORIZATION' protocol and a 'FORBIDDEN LIST' of AI defaults, which are techniques used to force the model into a specific operational mode beyond standard system prompt instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting and auditing untrusted project code provided by the user (referenced in 'SKILL.md' and processed by 'scripts/mobile_audit.py'). This creates an attack surface where malicious instructions embedded in the audited code could potentially influence the agent's behavior during the analysis phase. The skill lacks explicit sanitization or instruction-isolation markers for the content it reads.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a provided Python script ('scripts/mobile_audit.py') against a user-specified directory. Although the script performs local static analysis using standard libraries, the instruction for the agent to 'Execute these for validation (don't read, just run)' is a pattern that encourages running potentially complex code without prior review.
Audit Metadata