executing-plans

SUSPICIOUS: the core behavior matches the stated purpose of executing a saved development plan, but it also executes plan-derived `verifyCommand` content and depends on other skills, creating meaningful command-execution and transitive-trust risk. No direct credential harvesting, exfiltration endpoint, or malicious installer behavior is present in this skill text.