lfg

SUSPICIOUS: the skill's capabilities broadly match its purpose, but it enables autonomous external actions (commit/push/PR/CI handling) and relies on transitive child-skill trust. No direct malware or credential-harvesting behavior is evident in this skill alone, but the autonomy level makes it high operational risk.