The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill provides the run-code and eval commands, enabling the execution of arbitrary JavaScript within the browser context. This functionality is used for geolocation emulation, permission overrides, and custom scraping logic.
Evidence: Commands such as playwright-cli run-code "async page => ..." and playwright-cli eval "document.title" are documented in SKILL.md and references/running-code.md.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to browse and process content from external websites, creating a surface for instructions embedded in web pages to influence agent behavior.
Ingestion points: playwright-cli goto, playwright-cli snapshot, and playwright-cli find (documented in SKILL.md).
Boundary markers: No explicit delimiters or instructions to ignore embedded content are implemented.
Capability inventory: Includes file system writes (screenshot, pdf, state-save), network request interception/mocking, and JavaScript execution.
Sanitization: No sanitization of ingested HTML or text content is described.
[DATA_EXFILTRATION]: The skill allows for the retrieval and storage of sensitive browser data, including session cookies and local storage content.
Evidence: playwright-cli state-save, playwright-cli cookie-list, and playwright-cli localstorage-get commands in SKILL.md and references/storage-state.md.
[EXTERNAL_DOWNLOADS]: The shell wrapper script fetches the @playwright/cli package from Microsoft's official npm registry if the global binary is unavailable.
Evidence: scripts/playwright_cli.sh contains the command npx --yes --package @playwright/cli@latest playwright-cli.

playwright-cli