shadcn
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the shadcn CLI through runners like npx to search for and install UI components. These actions are standard for the tool's intended use in development environments.\n- [DYNAMIC_CONTEXT_INJECTION]: The SKILL.md file uses the dynamic context syntax (!) to run 'npx shadcn@latest info --json' at load time. This is used to gather project-specific metadata such as framework type and import aliases for the AI agent.\n- [EXTERNAL_DOWNLOADS]: The skill allows the agent to fetch components from official, community, and arbitrary GitHub registries. It includes specific workflow steps for the agent to dry-run, diff, and audit all added code to ensure it meets security and project standards.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests component data from third-party sources, creating an attack surface for instructions embedded in component code. This is mitigated by explicit instructions for the agent to manually review and verify all external files.\n
- Ingestion points: Installation of items from registries via the CLI.\n
- Boundary markers: Absent for prompted data, but the CLI supports diffing for manual review.\n
- Capability inventory: Writing files to the project directory and performing network requests via the CLI.\n
- Sanitization: The skill mandates a verification process where the agent must read and verify files after they are added.
Audit Metadata