skills/secondsky/claude-skills/zod/Gen Agent Trust Hub

zod

Pass

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns detected. The skill primarily serves as a documentation resource for schema validation with Zod.
  • [METADATA_POISONING]: The skill documentation refers to 'Zod 4.x' and includes features like z.codec() and z.iso.* that do not exist in the current stable version of the library (v3.23.x). Additionally, the last_verified date is set in the future (2025-11-17). While these discrepancies result in non-functional code examples for the current version of Zod, they do not facilitate a security breach or direct users toward malicious third-party scripts or registries.
  • [EXTERNAL_DOWNLOADS]: The skill suggests installing the standard zod package via legitimate package managers (bun, yarn, npm) from the official registry. All referenced ecosystem packages (e.g., trpc, react-hook-form, hono) are well-known and legitimate.
  • [REMOTE_CODE_EXECUTION]: No remote code execution patterns, such as piping network downloads to a shell or using eval() on untrusted input, were found in the skill or its references.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 9, 2026, 02:16 PM