The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/with_server.py utility is designed to execute arbitrary shell commands provided via the --server command-line argument. This allows the agent to start local development servers (e.g., 'npm run dev') before running automation scripts.
[DYNAMIC_EXECUTION]: The script scripts/with_server.py utilizes subprocess.Popen with shell=True. While this enables the use of shell features like command chaining ('&&') and directory changes ('cd'), it represents a security risk if the command strings are constructed from untrusted input.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the scraping and inspection of web applications through Playwright commands such as page.content() and page.locator().all(). This creates an attack surface where a malicious web application could embed hidden instructions to manipulate the agent.
Ingestion points: Web page content and DOM elements are ingested into the agent's context during reconnaissance steps defined in SKILL.md.
Boundary markers: No explicit boundary markers or instructions to ignore embedded content are provided in the example scripts.
Capability inventory: The skill possesses shell execution capabilities via scripts/with_server.py and the ability to interact with the file system.
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from the browser before it is processed by the agent.

webapp-testing