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. It also uses subprocess.run to execute the final automation command. While this is the intended functionality for managing local development environments, it grants the agent the ability to execute arbitrary shell commands.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted external data from web pages through Playwright's browser automation.
Ingestion points: The SKILL.md recommends using page.content() to inspect the DOM, and examples/console_logging.py captures and stores browser console messages using page.on("console", ...).
Boundary markers: None are present in the documentation or scripts to delimit external web content from agent instructions.
Capability inventory: The skill possesses capabilities for file writing (examples/console_logging.py), screenshot capture (examples/element_discovery.py), and arbitrary shell command execution (scripts/with_server.py).
Sanitization: No sanitization or filtering is performed on the captured HTML content or console logs before they are processed by the agent or written to disk.

webapp-testing