skills/jimliu/open-artifact/slides/Gen Agent Trust Hub

slides

Pass

Audited by Gen Agent Trust Hub on Nov 21, 2026

Risk Level: SAFEPERSISTENCEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PERSISTENCE]: The skill instructs the agent to write a .git/hooks/pre-commit shell script in the user's repository. This hook is designed to automatically execute the skill's CLI script (scripts/cli.mjs) whenever a git commit is initiated, ensuring the skill's logic runs persistently during development workflows.
  • [INDIRECT_PROMPT_INJECTION]: The skill operates by reading and revising locally stored HTML and JSON files which may originate from untrusted sources. Although the instructions include explicit warnings for the agent to treat this content as data and not as instructions, the ingestion of user-controlled markup represents an attack surface for indirect prompt injection.
  • [EXTERNAL_DOWNLOADS]: The skill references and downloads web font resources from Google Fonts (fonts.googleapis.com). This is documented as a well-known and trusted service, but it involves fetching external assets at runtime.
  • [COMMAND_EXECUTION]: The skill relies extensively on executing a local Node.js script (scripts/cli.mjs) to perform deck validation, server management, and file operations. It also utilizes standard system commands such as cp for file management during design system installation.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 21, 2026, 02:54 AM