The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/smoke-test.mjs uses createRequire to dynamically load the playwright module from a path computed at runtime using the current working directory. This pattern allows the utility to adapt to different project environments but qualifies as dynamic loading from a computed path.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content by navigating a headless browser to a user-provided URL or local file and reporting the resulting console errors and exceptions. This creates a surface where a malicious page could inject instructions into the reported logs to influence the AI agent.\n
Ingestion points: The target argument in scripts/smoke-test.mjs which accepts arbitrary URLs and file paths.\n
Boundary markers: None; the output from the browser is not wrapped in markers to distinguish it from the skill's instructions.\n
Capability inventory: JavaScript execution via Playwright and monitoring of all console output and runtime exceptions.\n
Sanitization: The script does not perform any sanitization or escaping of the messages captured from the browser environment before reporting them to the user or agent.

smoke-testing-web-games