dotnet-design-pattern-review

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted source code provided by the user via the ${selection} variable.
  • Ingestion points: The ${selection} placeholder in SKILL.md is the primary entry point for external data.
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to help distinguish between the skill's instructions and potentially malicious instructions embedded within the code being reviewed.
  • Capability inventory: While the skill's instructions explicitly forbid code modifications ("Do not make any changes to the code, just provide a review"), the YAML frontmatter does not utilize the allowed-tools field to restrict the agent's default toolset, which may include file access or command execution depending on the host environment.
  • Sanitization: There is no evidence of input validation or sanitization for the content of the ${selection} variable.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 11:19 PM
Security Audit — agent-trust-hub — dotnet-design-pattern-review