The Agent Skills Directory

[SAFE]: The skill provides instructions and templates for generating Playwright end-to-end tests. It correctly identifies the need for exploring the codebase to match existing conventions (TypeScript vs JavaScript, page objects, fixtures).
[COMMAND_EXECUTION]: The skill includes a step to verify the generated tests using npx playwright test. This is a standard and expected behavior for a test generation tool. The command targets the specifically generated file, minimizing risk.
[CREDENTIALS_SAFE]: While the skill handles authentication flows, it uses placeholder values like user@example.com and password123 in its templates. It also correctly advises checking for existing auth.setup.ts or storageState configuration, which are standard Playwright patterns for managing authentication state securely.
[DATA_EXPOSURE]: The skill reads project configuration (playwright.config.ts) and existing tests to ensure generated code matches the environment. This is local file access necessary for its primary purpose.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user descriptions or file paths via $ARGUMENTS. It applies strict rules for locator priority (e.g., getByRole, getByLabel) and web-first assertions, which acts as a structural boundary against malicious content influencing the generated code beyond the intended test logic.

generate