The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to download and run the @playwright/mcp package from the NPM registry. This package is an official component from Microsoft, which is a trusted organization. The workflow is documented in SKILL.md and usage-patterns.md.
[DYNAMIC_EXECUTION]: As part of its browser automation capabilities, the skill exposes a browser_run_code operation. This allows for the execution of arbitrary JavaScript within the controlled browser environment, as seen in usage-patterns.md.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data by navigating to and snapshotting websites using browser_navigate and browser_snapshot. This creates a vulnerability where instructions embedded in a webpage could be executed by the agent.
Ingestion points: Web content retrieved via browser navigation as described in SKILL.md and usage-patterns.md.
Boundary markers: The instructions do not define specific delimiters or warnings to ignore instructions found within the processed web data.
Capability inventory: The skill allows for navigation, DOM interaction, and script execution (browser_run_code) inside the browser context.
Sanitization: There are no documented mechanisms for sanitizing or filtering content from web pages before it is processed by the agent.

playwright-mcp-skill