shadcn

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). Several URLs point to custom or personal-hosted registries/JSON endpoints (api.npoint.io and templated private/acme registry URLs) which can deliver arbitrary code or component manifests outside official registries and are commonly used to distribute malicious payloads if untrusted, so they should be treated as suspicious.

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

• Third-party content exposure detected (low risk: 0.10). The required runtime workflow primarily operates on first-party injected project context (e.g., npx shadcn@latest info / components.json) and only reads outsider-authored free text when explicitly instructed to fetch/view registry item content via CLI/MCP tools (e.g., docs, view, search that then fetches registry files and example code), which is actively searchable/selectable rather than continuously monitored.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent to run the shadcn CLI and to fetch external docs/presets at runtime (e.g. https://ui.shadcn.com and the example URL https://api.npoint.io/abc123), which will retrieve remote content that the agent is expected to ingest or that the CLI may execute/install — therefore these URLs are runtime external dependencies that can control prompts or execute code.