design-compare

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze external references, such as URLs or competitor products, which introduces a surface for indirect prompt injection.
  • Ingestion points: The workflow in SKILL.md includes an 'Analyze reference (from URL, description, or known product)' step where untrusted external data enters the agent's context.
  • Boundary markers: The instructions lack specific guidance for the agent to treat external content as data rather than instructions, and no delimiters are specified.
  • Capability inventory: The skill requires file-reading capabilities ('Audit current project design') and likely web-browsing capabilities to retrieve external references.
  • Sanitization: There are no defined mechanisms for sanitizing or validating the content retrieved from external URLs before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 02:48 PM
Security Audit — agent-trust-hub — design-compare