ios-automate

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill provides technical instructions for interacting with an iOS simulator via the axe command-line tool. All operations described are standard automation tasks such as tapping, swiping, and typing within the context of a developer's simulator environment.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection attack surface as it instructs the agent to parse and interpret UI element data (such as labels and text) retrieved from external mobile applications using the axe describe-ui command.
  • Ingestion points: Text and labels retrieved from the simulator UI via axe describe-ui in Step 1 and Step 6.
  • Boundary markers: Absent; there are no instructions to use delimiters or ignore embedded commands within the retrieved UI content.
  • Capability inventory: The agent has the ability to execute shell commands to interact with the simulator (tap, swipe, type, batch) and write to the file system (screenshot, record-video).
  • Sanitization: No sanitization or validation of the UI content is recommended; the agent relies on raw accessibility labels which could be controlled by a malicious application.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 01:51 AM
Security Audit — agent-trust-hub — ios-automate