The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to download and execute the punchlist-axi package from the public NPM registry. This package is an official resource provided by the skill author.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to start and manage a background proxy. Specifically, it uses $ARGUMENTS directly in a shell command (--target <port|url>), which creates a risk of command injection if the input contains shell metacharacters and is not properly sanitized before execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes HTML, CSS, and selector data captured from a browser via an element picker. This external content is untrusted and could be crafted to include malicious instructions that influence the agent's code-editing tasks.
Ingestion points: Browser data (HTML, styles, source-file hints) received through the proxy.
Boundary markers: No explicit delimiters or warnings are used when processing the browser data.
Capability inventory: The agent has the capability to write to the local file system (editing code) and execute shell commands.
Sanitization: The instructions do not specify any sanitization or validation of the data received from the browser element picker.

punchlist