shadcn-svelte
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
npx shadcn-svelte@latestto initialize projects and add components. Instructions include the-yflag to bypass user confirmation and the-oflag to overwrite existing files. - [EXTERNAL_DOWNLOADS]: Fetches component registries, documentation, and metadata from several external domains including
shadcn-svelte.com,bits-ui.com,formsnap.dev, andpaneforge.com. These are the official sites for the libraries described. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it instructs the agent to fetch a machine-readable registry from a remote URL and use the parsed 'name' fields directly in shell commands.
- Ingestion points: Remote JSON registry at
https://shadcn-svelte.com/registry/index.json. - Boundary markers: None present; the agent is told to use the live registry data directly.
- Capability inventory: Shell command execution via the
npxtool. - Sanitization: The instructions do not specify any validation or sanitization of the component names retrieved from the external registry before they are passed to the terminal.
Audit Metadata