vercel-composition-patterns

Warn

Audited by Gen Agent Trust Hub on Oct 6, 2026

Risk Level: MEDIUMMETADATA_POISONINGNO_CODESAFE
Full Analysis
  • [METADATA_POISONING]: The skill contains metadata in the SKILL.md file identifying the author as 'vercel'. This is an identity discrepancy as the skill is attributed to 'huangzida'. Such deceptive metadata can mislead users or automated systems regarding the skill's origin and safety.
  • [NO_CODE]: The repository consists entirely of Markdown documentation and code examples. There are no executable scripts, shell commands, or binary files that could pose a direct runtime threat.
  • [SAFE]: No suspicious network operations or unauthorized external downloads were detected. All external links point to official documentation sites.
  • [SAFE]: The skill does not attempt to access sensitive files, environment variables, or hardcoded credentials.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 6, 2026, 04:20 PM