e2e-test-builder
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides standard E2E testing patterns and best practices. Hardcoded strings such as 'admin@example.com' and 'password123' are clearly used as example test data within the context of flow demonstrations and do not represent leaked production secrets.\n- [COMMAND_EXECUTION]: The 'playwright.config.ts' example includes a 'webServer' block that executes 'npm run dev'. This is a standard and expected pattern in Playwright to ensure the application under test is running before execution begins.\n- [INDIRECT_PROMPT_INJECTION]: The skill provides a structural framework for generating code based on user-defined flows, which represents a potential ingestion point for instructions.\n
- Ingestion points: User-provided descriptions of critical flows and page interactions used to generate tests.\n
- Boundary markers: None explicitly defined in the provided snippets to isolate instructions from data.\n
- Capability inventory: Browser automation capabilities including navigation (page.goto), interaction (page.click, page.fill), and network mocking (page.route).\n
- Sanitization: None present in the templates as the skill serves as a static documentation guide.
Audit Metadata