build

SUSPICIOUS: the core purpose is coherent, but the skill has a moderately high operational footprint: autonomous subagents, shell execution, browser interaction with untrusted content, and GitHub push/PR actions. No direct exfiltration or malicious endpoint is shown, yet unresolved provenance for referenced skills and the high-privilege review/execution loop make it medium risk rather than benign.