skills/shotaiuchi/uiai/uiai-ios-test/Gen Agent Trust Hub

uiai-ios-test

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it interprets and executes UI actions based on external YAML scenario files.
  • Ingestion points: Scenario files provided through the scenarios argument as defined in SKILL.md.
  • Boundary markers: Test scenarios are executed as isolated sub-agents using the Task tool to prevent context contamination.
  • Capability inventory: Capabilities include executing simulator commands and UI manipulation tools (idb, simctl) as documented in the references folder.
  • Sanitization: The skill logic includes a pre-execution phase to parse variables and validate the environment as described in references/execution-flow.md.
  • [COMMAND_EXECUTION]: The skill executes local shell commands using xcrun simctl and idb to manage apps and simulators. These are standard operations for its intended function in iOS development.
  • [EXTERNAL_DOWNLOADS]: The documentation references the installation of fb-idb and idb-companion, which are official developer tools from a well-known service.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 07:14 PM
Security Audit — agent-trust-hub — uiai-ios-test