marketplace-scaffold

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • External Resource Fetching: The skill utilizes npx shadcn@latest add to fetch application components and registry configurations from marketplace-sdk.sitecorecloud.io. This is a standard method for distributing official SDK components and UI themes from well-known service providers.
  • Command Execution for Scaffolding: It provides commands to initialize Next.js projects and install SDK packages via npx. These operations are core to the skill's purpose of setting up a new development environment and are expected behaviors for a scaffolding tool.
  • Input Processing Considerations: The skill incorporates user-provided inputs, such as the application name, into shell commands (e.g., npx create-next-app). While this follows standard project creation patterns, it represents a surface where untrusted data influences command execution.
  • Ingestion points: User-provided app-name argument and architecture choices in SKILL.md.
  • Boundary markers: No explicit delimiters are used for user input interpolation.
  • Capability inventory: Subprocess calls via npx are present across the SKILL.md file.
  • Sanitization: The skill relies on the validation built into the underlying CLI tools (create-next-app, shadcn) to handle malformed input.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 10:11 AM
Security Audit — agent-trust-hub — marketplace-scaffold