skills/metamask/skills/e2e-test/Gen Agent Trust Hub

e2e-test

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the user or agent to execute local shell commands for the development lifecycle. This includes running yarn lint for code quality, yarn test:e2e:* for building the mobile application, and detox test for executing regression and smoke tests on simulators or emulators.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze existing codebase patterns (spec files, page objects, and selectors) to generate new tests. This creates a surface where instructions embedded in the analyzed code could influence the agent's behavior. However, the skill provides strict templates and 'Golden Rules' (e.g., mandatory use of withFixtures and Page Object Model) that serve as structural constraints.
  • [SAFE]: External URL references (e.g., metamask.io) and API mocking examples point to the vendor's own infrastructure or use standard placeholders (api.example.com). All code patterns and commands are standard for a professional React Native/Detox development environment and align with the stated purpose of the MetaMask Mobile repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 07:58 PM
Security Audit — agent-trust-hub — e2e-test