mobilewright

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes various CLI commands using npx for project initialization, environment checks, and running tests (e.g., npx mobilewright test, npx mobilewright doctor). It also performs automated searches within local project files like *.pbxproj, build.gradle, and app.json to identify application bundle IDs.
  • [EXTERNAL_DOWNLOADS]: The skill references and downloads the mobilewright framework and @mobilewright/test package from the official mobile-next vendor repository on GitHub.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the mobile application's UI hierarchy via the screen.viewTree() method.
  • Ingestion points: Data is retrieved from the live mobile device UI tree in SKILL.md.
  • Boundary markers: None present to distinguish UI content from instructions.
  • Capability inventory: The skill has capabilities for shell command execution (npx), file system reads, and network interaction via the framework.
  • Sanitization: There is no explicit sanitization or filtering of the UI hierarchy data before it is used by the agent to determine locators or verify implementation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 12:50 AM
Security Audit — agent-trust-hub — mobilewright