The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The CI workflow defined in references/ci-automation.md downloads Node.js, Playwright, and project dependencies using standard GitHub Actions and npm commands. These resources originate from well-known registries and official repositories.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from live websites during the testing process (e.g., through page.goto calls and network request monitoring in references/gdpr-cmp-tests.md). While this constitutes an attack surface, the risk is inherent to the skill's primary purpose of compliance auditing. No specific boundary markers or sanitization logic is implemented for the external content processed.
[DYNAMIC_EXECUTION]: The skill utilizes Playwright's page.evaluate() and page.addInitScript() methods in references/gdpr-cmp-tests.md and references/better-ads-tests.md to execute JavaScript logic within the browser context. This is a standard requirement for verifying client-side consent signals such as TCF v2 APIs and Google Consent Mode.
[COMMAND_EXECUTION]: The scheduled audit workflow in references/ci-automation.md involves local command execution for environment setup and test suite invocation, consistent with standard automated testing practices.

compliance-testing