The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documents the installation of standard tools and libraries using npm and npx, such as @playwright/test, @axe-core/playwright, and dotenv. These are well-known and trusted ecosystem packages for the Playwright framework.
[INDIRECT_PROMPT_INJECTION]: The skill explicitly identifies and warns against indirect prompt injection risks in SKILL.md and core/SKILL.md. It instructs developers to treat all returned page content as untrusted input and to never pass raw text back into agent instructions without sanitization.
[SECURITY_TESTING]: In core/security-testing.md, the guide provides common XSS, CSRF, and template injection payloads. These are included strictly for the purpose of testing the security of an application under development and do not target the AI agent or its environment.
[CLI_DOCUMENTATION]: The playwright-cli guides explain how to use the command-line interface for browser automation. Features like run-code and eval are documented as standard capabilities for interacting with the browser's DOM and context during automated tasks.
[BEST_PRACTICES]: The skill consistently recommends security best practices, such as using .env files for secrets, excluding authentication state from version control via .gitignore, and pinning external dependencies to immutable references in CI/CD pipelines.

playwright-skill