start-an-app

SUSPICIOUS. The stated purpose is coherent for an app scaffolding skill, but its trust footprint is too broad: it globally installs six transitive skills from mutable upstream sources before any build work, including at least one upstream skill with load-time execution of an unpinned `npx ...@latest` command. This is not confirmed malware, but it is a high supply-chain and transitive-trust risk for a skill whose core job could be done without expanding the agent's permission boundary this aggressively.