shadcn-ui

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs using the CLI to fetch and preview components from remote registries and arbitrary URLs (e.g., pnpx shadcn@latest add https://example.com/r/custom-component.json, pnpx shadcn@latest view <component-name>, and registry entries in components.json), so the agent will ingest public third-party content that could influence subsequent actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill instructs running the remote CLI (pnpx shadcn@latest add) which fetches and executes remote package code and component JSONs from registries (e.g., https://example.com/r/custom-component.json or https://registry.acme.com/r/{name}.json), so external content is fetched at runtime and can supply/execute code that controls the project.