playwright-mcp-setup

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly configures Claude to control a browser (steps 5–6 and the "Test the setup" example) and to use tools like browser_navigate and browser_evaluate to load and interact with arbitrary public URLs (e.g., "Take a screenshot of https://example.com"), which exposes the agent to untrusted third-party web content that can influence its actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes the external package via "npx @playwright/mcp@latest" (the @playwright/mcp npm package), which will fetch and execute remote code at runtime and is required for the MCP server to operate.