shadcn

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md and CLI docs explicitly instruct the agent to run commands like "npx shadcn@latest docs" and to fetch the resolved documentation/example URLs (including public registry and raw GitHub URLs) and use MCP/view tools to read registry item contents, meaning the agent will ingest and act on untrusted third‑party, user-provided web content as part of its workflow.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs fetching component docs/examples at runtime (for example: https://raw.githubusercontent.com/...) via the CLI (npx shadcn@latest docs / add) and uses that fetched content to drive the agent's guidance and code changes, so remote raw GitHub URLs can be injected into the agent context and directly control instructions.