The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill fetches and executes the @playwright/cli package from the official npm registry using npx. This ensures that browser automation tools are available in the runtime environment. Since the package originates from a trusted organization, this is documented as a standard functional dependency.
[COMMAND_EXECUTION]: A shell wrapper script (scripts/playwright_cli.sh) is provided to execute various browser automation commands, including script evaluation and interaction commands like clicking and typing.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection. * Ingestion points: The agent ingests external web content via snapshot (DOM structure) and eval (text extraction) commands as seen in SKILL.md and references/cli.md. * Boundary markers: There are no explicit delimiters or instructions provided to isolate potentially untrusted web content from the agent's core instructions. * Capability inventory: The skill possesses significant capabilities, including the ability to fill forms, click buttons, and run arbitrary JavaScript within the browser session. * Sanitization: No sanitization or validation of the data retrieved from external web pages is performed before the agent processes it.

playwright