The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute shell commands using the npx podbenchmark utility.
[EXTERNAL_DOWNLOADS]: The skill fetches the podbenchmark package from the npm registry at runtime using npx.
[DYNAMIC_EXECUTION]: The skill utilizes npx to dynamically fetch and execute code, which can vary across executions.
[INDIRECT_PROMPT_INJECTION]: The skill processes external supplier links, which serves as an ingestion point for untrusted data.
Ingestion points: Supplier and product URLs passed as arguments to the compare command in SKILL.md.
Boundary markers: The instructions do not define delimiters or provide warnings to ignore potential instructions embedded in the external content.
Capability inventory: Shell command execution via the npx environment.
Sanitization: There is no description of logic to sanitize or validate the content retrieved from the external links.

podbenchmark