axe-ios-simulator

Pass

Audited by Gen Agent Trust Hub on Aug 9, 2026

Risk Level: SAFENO_CODEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the axe tool from a third-party Homebrew tap (cameroncooke/axe/axe). While this is an external dependency, it is standard for this type of utility.
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to interact with the iOS simulator, performing actions such as tapping, typing, and hardware button simulation. These operations are within the stated purpose of the skill.
  • [DATA_EXFILTRATION]: The skill includes capabilities to capture screenshots (axe screenshot) and record video (axe record-video) of the simulator. Users should exercise caution when automating apps that display sensitive information, as these media captures could expose that data.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) because it processes UI metadata from the simulator to drive automation.
  • Ingestion points: UI metadata (accessibility labels, identifiers) is ingested via axe describe-ui in the SKILL.md file's automation patterns.
  • Boundary markers: No explicit delimiters or warnings are provided to the agent to treat UI metadata as untrusted.
  • Capability inventory: The skill possesses the ability to execute shell commands and perform file writes (screenshots).
  • Sanitization: There is no evidence of sanitization of UI labels before they are used in control flow logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 9, 2026, 05:07 PM
Security Audit — agent-trust-hub — axe-ios-simulator