klimkit-web-design-guidelines

Warn

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: MEDIUMPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: Deceptive metadata identified. The skill's author is listed as 'vercel' in the YAML frontmatter, but the system identifies the actual author as 'klimentij'. This impersonation of a well-known organization can mislead users regarding the skill's origin and safety.
  • [EXTERNAL_DOWNLOADS]: The skill fetches dynamic guidelines from the Vercel Labs repository on GitHub. While the source is a trusted organization, the fetched content serves as the logic for the audit, effectively implementing remote instruction loading.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted user code/files for review.
  • Ingestion points: User-specified files or file patterns in the review command.
  • Boundary markers: Absent; there are no instructions to the agent to treat file content exclusively as data or to ignore embedded instructions.
  • Capability inventory: File system read access and network fetching via WebFetch.
  • Sanitization: None; the skill does not specify any pre-processing or validation of the input file content.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 19, 2026, 07:34 PM
Security Audit — agent-trust-hub — klimkit-web-design-guidelines