The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads the @playwright/cli package from the official NPM registry using npx in the scripts/playwright_cli.sh wrapper script. This is the standard method for executing this utility and originates from a trusted organization.
[COMMAND_EXECUTION]: A local bash script (scripts/playwright_cli.sh) is used to execute the Playwright CLI for tasks such as clicking, typing, and capturing page states.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes content from untrusted external websites.
Ingestion points: Web page content is ingested through snapshot and eval commands as described in SKILL.md and references/workflows.md.
Boundary markers: The skill does not provide specific delimiters or instructions to treat web content as data rather than instructions.
Capability inventory: The agent can navigate, fill forms, and execute arbitrary JavaScript code via eval and run-code within the browser context, as listed in references/cli.md.
Sanitization: There is no evidence of filtering or validation of the content retrieved from web pages before it is processed by the agent.

playwright