The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to dynamically download and execute the @playwright/mcp and playwright-cli packages from the NPM registry at runtime.
[COMMAND_EXECUTION]: Provides a wrapper script (scripts/playwright_cli.sh) that executes shell commands to drive browser interactions via the Playwright CLI.
[REMOTE_CODE_EXECUTION]: Includes eval and run-code capabilities which allow the execution of arbitrary JavaScript within the context of the automated browser instance.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from external websites.
Ingestion points: Browser page content is ingested into the agent's context through the snapshot command and command-line arguments.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands in web content are provided in the skill definitions.
Capability inventory: The skill can perform network operations (navigation), file system writes (screenshots, traces), and arbitrary code execution in the browser context (eval).
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from websites before it is processed by the agent.

playwright