The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The script scripts/ports.sh uses npx --yes port-whisperer to fetch the port-whisperer package from the public npm registry if it is not already available on the system PATH.
[REMOTE_CODE_EXECUTION]: By using npx --yes, the skill performs automatic execution of code downloaded from a remote package registry without version pinning or integrity verification at the script level.
[COMMAND_EXECUTION]: The skill provides the ability to execute system-level commands through the port-whisperer CLI, including kill and clean actions that can terminate running processes. The wrapper script scripts/ports.sh passes user-supplied arguments directly to these commands.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from the local system (process names, command-line arguments, and application logs) which could potentially contain malicious instructions designed to influence the agent's subsequent actions.
Ingestion points: Outputs from list, ps, inspect, and logs commands in scripts/ports.sh.
Boundary markers: The skill relies on natural language safety rules in SKILL.md but does not implement technical boundary markers or 'ignore' instructions in the data-processing scripts.
Capability inventory: The skill has access to the Bash tool to execute scripts/ports.sh, which can terminate system processes.
Sanitization: The script includes a strip_ansi function to remove color codes but does not perform content sanitization or validation of the process data before it is returned to the agent context.

port-whisperer