claude-design-system-hooks

SUSPICIOUS. The stated design-to-code purpose is plausible, but the actual install and auth flow relies on unverifiable third-party packages/CLI and a container image that do not match Anthropic's documented Claude Design setup. Because the skill asks users to pass API keys into that unverified tooling and can install further community packages, its footprint is disproportionate and high risk even without confirmed malicious payloads.