frontend-dev-guidelines

Warn

Audited by Gen Agent Trust Hub on Apr 24, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends several non-standard Node.js packages, including react-hook-blog and @hookblog/resolvers (found in resources/complete-examples.md). These appear to be corrupted versions of well-known libraries like react-hook-form and @hookform/resolvers. Recommending unknown or non-existent packages from public registries presents a significant supply chain risk, as an attacker could register these names to deliver malicious code.
  • [REMOTE_CODE_EXECUTION]: If an AI agent or developer follows the instructions to install the unverified react-hook-blog package, it could lead to remote code execution if the package on the registry contains malicious post-install scripts or logic.
  • [PROMPT_INJECTION]: The skill contains pervasive and systematic text corruption where technical terms have been modified (e.g., 'performance' becomes perblogance, 'transformers' becomes transblogers, and the HTML <form> tag is replaced with <blog>). This pattern of 'blog' for 'form' substitution is used to poison the agent's instructions, likely resulting in the generation of non-functional or intentionally broken code that uses incorrect syntax and library names.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 24, 2026, 10:22 AM
Security Audit — agent-trust-hub — frontend-dev-guidelines