The Agent Skills Directory

[PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from the browser environment, creating a surface for indirect prompt injection.
Ingestion points: references/debugging-flaky.md includes instructions for the agent to listen to console and pageerror events from web pages being tested.
Boundary markers: The skill does not provide instructions to help the agent differentiate between its core instructions and data captured from external websites.
Capability inventory: The agent has the capability to write files (screenshots and traces), intercept network requests via API mocking, and execute shell commands.
Sanitization: There is no guidance provided for sanitizing or escaping the content captured from the browser logs before it is processed by the agent.
[COMMAND_EXECUTION]: The skill contains multiple examples and instructions for executing shell commands for test execution and environment setup.
Evidence: Commands such as npx playwright test, npx playwright show-trace, and npm run dev are documented across SKILL.md and references/configuration.md.
[EXTERNAL_DOWNLOADS]: The skill references the installation of required browser binaries and uses established external workflows for CI/CD integration.
Evidence: The configuration in references/configuration.md uses npx playwright install --with-deps and several official GitHub actions including actions/checkout, actions/setup-node, and actions/upload-artifact.

playwright-expert