The Agent Skills Directory

Local Command Execution: The scripts/with_server.py utility employs subprocess.Popen with the shell=True parameter to initiate local development servers. * Evidence: Found in the main function of scripts/with_server.py. * Context: This implementation allows the tool to handle multi-part shell commands (like directory changes or piping) frequently used in web development workflows. It is a functional requirement for starting arbitrary local servers within the testing context.
Interaction with External Web Content (Indirect Prompt Injection Surface): The skill utilizes the Playwright library to navigate and inspect web pages, which exposes an interface to external data. * Ingestion points: page.content() in examples/element_discovery.py and console message capture in examples/console_logging.py. * Boundary markers: None explicitly defined in the provided examples. * Capability inventory: Command execution via subprocess in scripts/with_server.py and file writing in example scripts. * Sanitization: External content is processed directly without specific sanitization steps. * Context: Processing content from external websites or application logs can occasionally lead to indirect prompt injection if the content contains instructions designed to influence the agent. This is a common property of browser automation tools and is typically managed by broader system security layers.

webapp-testing