do-execute-task

SUSPICIOUS: the skill’s core behavior matches its stated purpose, but its footprint is unusually forceful and high-trust. The main concern is not malware indicators or rogue credential routing; it is the mandatory no-confirmation autonomy combined with code execution, file modification, task-state updates, dev-server startup, and optional external-content ingestion through MCP/web search. Install trust is relatively normal, but operational risk is medium-high because the skill is designed to act broadly and immediately.