The Agent Skills Directory

[SAFE]: The skill serves as a technical guide for expert QA automation using Playwright. All provided code examples follow industry standard practices for end-to-end testing.
[DATA_EXPOSURE]: The instructions describe the use of storageState files (e.g., playwright/.auth/user.json) to persist authentication. While these files contain session data, their use is a standard Playwright feature. The skill correctly implies these are local configuration files.
[PROMPT_INJECTION]: As a tool designed to interact with external web pages via page.goto() and various locators, the skill possesses an inherent surface for indirect prompt injection (i.e., content from a website attempting to influence the agent).
Ingestion points: Browser navigation and interaction via page.goto and page.getByRole (SKILL.md).
Boundary markers: Absent; instructions do not explicitly warn the agent to ignore instructions embedded in the target website's content.
Capability inventory: The agent can perform network requests and execute local scripts via the webServer configuration (npm run dev).
Sanitization: Relies on the default Playwright browser sandbox and the LLM's internal safety guardrails.
Conclusion: This is a standard risk for any web automation tool and is not indicative of malicious intent within the skill itself.

Playwright Enhanced