The Agent Skills Directory

[COMMAND_EXECUTION]: The helper script scripts/with_server.py uses subprocess.Popen with shell=True to execute server startup commands provided via arguments. It also uses subprocess.run to execute automation scripts. This is intended functionality to support complex commands (e.g., using cd and &&) for local server management.
[EXTERNAL_DOWNLOADS]: The documentation recommends the use of expect-cli from the millionco/expect project for adversarial browser testing.
[INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection as it ingests untrusted data from web pages.
Ingestion points: Data enters the context via page.content(), page.locator().all(), and page.on("console", ...) in examples/element_discovery.py and examples/console_logging.py.
Boundary markers: The skill does not implement specific delimiters or 'ignore' instructions for the data extracted from browser sessions.
Capability inventory: The skill possesses capabilities for file system writes (saving screenshots and logs) and command execution (via scripts/with_server.py).
Sanitization: There is no explicit sanitization or validation of the content extracted from web pages before it is logged or processed by the agent.

webapp-testing