code-maintainability

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of markdown documentation providing best practices for architectural design and code readability. No malicious code, base64 obfuscation, or sensitive credential exposure were detected.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data in the form of code diffs and project files. While this creates a potential attack surface, the risk is inherent to the intended primary purpose of code review and is addressed through architectural guidelines.\n
  • Ingestion points: User-provided code diffs and source files analyzed during the code review process.\n
  • Boundary markers: No specific delimiters or safety warnings are established in the instructions to separate untrusted code from agent instructions.\n
  • Capability inventory: The skill is used in contexts where agents may have file system access for refactoring and authoring code.\n
  • Sanitization: No explicit content sanitization or escaping mechanisms are provided for handling input data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 01:16 AM
Security Audit — agent-trust-hub — code-maintainability