The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/with_server.py uses subprocess.Popen with shell=True to execute server startup commands provided via the --server argument. While this is intended to support complex shell operations like directory changes and command chaining (e.g., cd backend && python server.py), it poses a risk of command injection if the input strings are not strictly controlled.
[COMMAND_EXECUTION]: The scripts/with_server.py script executes an arbitrary test command (provided as trailing arguments) using subprocess.run, which allows the execution of any system command the agent specifies.
[PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from external or local web applications, creating a surface for indirect prompt injection.
Ingestion points: page.content(), button.inner_text(), and console logs captured in examples/element_discovery.py and examples/console_logging.py.
Boundary markers: The provided examples do not demonstrate the use of delimiters or 'ignore' instructions when processing extracted web content.
Capability inventory: The skill includes powerful capabilities such as arbitrary shell command execution via scripts/with_server.py.
Sanitization: No evidence of sanitization or validation of the extracted DOM content or logs is present in the example scripts before the data is processed or stored.

webapp-testing