web-artifacts-builder

Pass

Audited by Gen Agent Trust Hub on Oct 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell scripts (scripts/init-artifact.sh and scripts/bundle-artifact.sh) to automate the setup of a React development environment and the subsequent bundling of the project using Parcel.
  • [EXTERNAL_DOWNLOADS]: Fetches standard frontend development libraries and tools from the official NPM registry, including React, Vite, Tailwind CSS, and various shadcn/ui dependencies. It also checks for and optionally installs the pnpm package manager globally via npm.
  • [DYNAMIC_EXECUTION]: Programmatically generates configuration files for PostCSS, Tailwind CSS, and Vite, and uses Node.js to modify TypeScript configurations (tsconfig.json) at runtime to establish path aliases and project settings.
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the generation of code based on user-provided requirements, which introduces a potential attack surface for indirect prompt injection if malicious input is embedded into the generated artifact logic.
  • Ingestion points: User-provided specifications for artifact development (referenced in SKILL.md) are used to drive source code creation.
  • Boundary markers: No explicit delimiting or safety markers are defined to isolate user instructions within the code generation workflow.
  • Capability inventory: The skill maintains the ability to execute shell commands, install system-wide packages via NPM/PNPM, and perform local file system writes across multiple scripts.
  • Sanitization: No validation or sanitization mechanisms are present for user-supplied input before it is incorporated into the React application source.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 29, 2026, 09:44 AM