The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/with_server.py facilitates the execution of arbitrary shell commands provided by the user or agent. It specifically uses subprocess.Popen with shell=True to launch server processes and subprocess.run to execute the subsequent automation command. While this behavior is central to the skill's utility for managing development servers, it enables the execution of any system command passed to the helper script.
[INDIRECT_PROMPT_INJECTION]: The skill uses Playwright to fetch and process content from web pages, which represents a vulnerability surface where external data could influence agent behavior.
Ingestion points: In examples/element_discovery.py, examples/console_logging.py, and examples/static_html_automation.py, the agent reads page content using page.content(), page.locator().all(), and page.on("console", ...).
Boundary markers: The instructions and examples do not include boundary markers or warnings to the agent to ignore instructions embedded within the target web page's text or HTML comments.
Capability inventory: The skill has the capability to write to the filesystem (e.g., /mnt/user-data/outputs/console.log and screenshots in /tmp/) and execute shell commands through the with_server.py utility.
Sanitization: There is no evidence of sanitization or filtering applied to the DOM content or console logs before they are processed by the agent or written to files.
[DYNAMIC_EXECUTION]: The scripts/with_server.py file demonstrates dynamic execution patterns by using the subprocess module to run strings as shell commands, supporting complex command chains (e.g., using cd and &&).

webapp-testing