The Agent Skills Directory

[SAFE]: The skill is a legitimate educational and procedural resource for quality assurance engineering. It focuses on writing maintainable, production-grade Playwright tests.
[CREDENTIALS_UNSAFE]: The skill follows security best practices for secret management. It instructs agents to use environment variables for sensitive data (e.g., process.env.TEST_USER_PASSWORD) and explicitly advises users to add the .auth/ directory to .gitignore to prevent session token leakage.
[DATA_EXFILTRATION]: No unauthorized network operations were found. The examples provided for network requests are limited to mocking the application under test's API or sending status updates to a user-configured Slack webhook in a reporting example.
[REMOTE_CODE_EXECUTION]: All code samples are standard TypeScript and Playwright API calls. There are no instances of downloading and executing remote scripts or using unsafe functions like eval().
[PROMPT_INJECTION]: The instructions reinforce agent safety and adherence to best practices by providing a "Common AI Agent Mistakes" list, which acts as a set of negative constraints to improve output quality rather than bypassing safety filters.

playwright-automation