shadcn

Pass

Audited by Gen Agent Trust Hub on Oct 11, 2026

Risk Level: SAFEDYNAMIC_CONTEXT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_CONTEXT_INJECTION]: The skill uses the dynamic context syntax in SKILL.md to execute npx shadcn@latest info --json when the skill is loaded. This is used to automatically populate the agent's context with project-specific settings, import aliases, and a list of installed components.
  • [COMMAND_EXECUTION]: The skill relies on shell commands (npx, pnpm, bun) to interact with the shadcn CLI. These commands allow the agent to add components, search registries, and initialize projects. This is the core intended functionality for a developer-oriented skill.
  • [EXTERNAL_DOWNLOADS]: The skill fetches the shadcn CLI from the npm registry and retrieves component documentation and example code from ui.shadcn.com and GitHub repositories. It also installs typography packages from the official Fontsource registry.
  • [INDIRECT_PROMPT_INJECTION]: The design-system.md workflow involves processing untrusted external data from a DESIGN.md file or brand specification.
  • Ingestion points: The agent reads the content of a user-provided DESIGN.md file.
  • Boundary markers: No specific boundary markers are used in the instructions for this ingestion.
  • Capability inventory: The skill has capabilities to modify the file system (component source code, global CSS) and execute shell commands via the CLI.
  • Sanitization: The instructions provide a structured workflow for extracting specific tokens (colors, typography), which acts as a functional filter for the content.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 11, 2026, 05:12 PM