AGENT LAB: SKILLS

nextjs-shadcn-builder

Warn

Audited by Snyk on Feb 15, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.70). The skill explicitly integrates with and queries the public shadcn MCP/docs server (e.g., https://ui.shadcn.com/docs/mcp and via "npx shadcn... mcp init") and directs the agent to read and use those external, public documentation responses as part of its component-discovery workflow, exposing it to untrusted third-party content.

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

  • Potentially malicious external URL detected (high risk: 0.80). The skill explicitly requires runtime queries to the shadcn MCP server (e.g., https://ui.shadcn.com/docs/mcp) for component discovery and examples, meaning remote content fetched during execution would directly inform agent prompts/instructions and is a required dependency.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 15, 2026, 08:15 PM