checking-idioms-and-consistency

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is entirely composed of markdown instructions and metadata. It does not contain any executable scripts, binaries, or active code components that could perform malicious actions.\n- [PROMPT_INJECTION]: The skill processes untrusted code diffs, which represents an indirect prompt injection attack surface (Category 8).\n
  • Ingestion points: The skill is designed to ingest and analyze code diffs provided in the agent's context (SKILL.md).\n
  • Boundary markers: The instructions do not define specific delimiters or instructions to treat the diff content as untrusted data or to ignore embedded instructions within it.\n
  • Capability inventory: The skill is purely analytical and provides text-based feedback; it does not request or use tools with the capability to perform network requests, execute code, or modify sensitive system files.\n
  • Sanitization: No sanitization or validation mechanisms are described for the incoming code diff content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 11:24 PM
Security Audit — agent-trust-hub — checking-idioms-and-consistency