vercel-composition-patterns

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists entirely of markdown files containing React architectural guidelines and code examples. Analysis of the instructions and referenced files shows no evidence of malicious patterns, network exfiltration, or unauthorized file access.
  • [NO_CODE]: The skill does not contain any functional scripts (Python, JavaScript, Shell) or automation logic. It serves as a static knowledge base for React development patterns to guide an agent's code generation and refactoring tasks.
  • [METADATA_POISONING]: The skill identifies 'vercel' as the author in its metadata, whereas the source context indicates the author is 'gigs-slc'. This discrepancy appears to be a common artifact of mirrored technical documentation rather than a deceptive attempt to escalate privileges or bypass safety protocols.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 10:18 AM