The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/lighthouse_audit.py uses the subprocess.run function to execute the lighthouse command-line tool for performance auditing. The implementation follows security best practices by passing arguments as a list and not using shell=True, which prevents command injection via the target URL.
[EXTERNAL_DOWNLOADS]: The script documentation notes a dependency on the well-known lighthouse CLI tool, which is expected for this functionality. The skill itself does not perform any remote code execution or unauthorized downloads.
[DATA_EXFILTRATION]: Analysis shows no attempts to access sensitive files (such as credentials or environment variables) or exfiltrate data to external servers. The script only processes the target URL and the resulting audit report.
[PROMPT_INJECTION]: The instructions in SKILL.md are strictly limited to performance optimization guidelines and do not contain any patterns designed to override agent behavior or bypass safety filters.

performance-profiling