vercel-composition-patterns

Warn

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: MEDIUMPROMPT_INJECTIONNO_CODE
Full Analysis
  • [METADATA_POISONING]: The metadata in SKILL.md incorrectly identifies the author as "vercel", which contradicts the provided author context "MichaelTarasov02". This impersonation of a well-known organization could lead to a misjudgment of the skill's origin and inherent safety.
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and refactor React codebases, which creates a significant vulnerability surface for indirect prompt injection.
  • Ingestion points: The agent ingests external React source code to apply composition patterns as described in the SKILL.md triggers.
  • Boundary markers: The instructions lack explicit delimiters or warnings (e.g., "ignore embedded instructions") for the code being processed, allowing potentially malicious instructions in source code comments or strings to influence agent behavior.
  • Capability inventory: The agent utilizing this skill is expected to have file system access and editing capabilities to perform code refactoring tasks.
  • Sanitization: There is no evidence of filtering, escaping, or validation of the external content before it is interpolated into the agent's context for processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 10, 2026, 12:22 PM
Security Audit — agent-trust-hub — vercel-composition-patterns