axe
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses shell commands like
axeandxcrun simctlto interact with the iOS simulator environment, including UI automation and hardware control. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by reading the accessibility tree of external iOS applications.
- Ingestion points: Data returned by the
axe describe-uicommand as described in SKILL.md. - Boundary markers: None identified; the agent is instructed to use the output to find elements without specific delimiters or isolation.
- Capability inventory: UI interaction (tapping, typing, swiping), screenshot capture to
/tmp, and simulator control (booting, shutting down). - Sanitization: No sanitization is mentioned for the content of accessibility labels, identifiers, or text fields, which could contain adversarial instructions intended to influence the agent.
Audit Metadata