ai-generated-business-code-review

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (MEDIUM): The skill's core function is to ingest and analyze untrusted external data (AI-generated business code). This creates a surface for indirect prompt injection.
  • Ingestion points: Processes external code provided in the agent's context for review.
  • Boundary markers: Absent. The instructions do not define delimiters (like triple backticks or XML tags) to isolate the code under review from the reviewer's logic.
  • Capability inventory: The skill generates scores, risk assessments, and checklists. While it does not execute code or access the network, its output directly influences human or automated decision-making regarding code quality and security.
  • Sanitization: No methods are specified to escape or ignore instructions embedded within the code comments or strings of the target file.
  • [Data Exfiltration] (SAFE): No evidence of network requests, file system access outside of standard context, or credential handling.
  • [Remote Code Execution] (SAFE): No patterns indicating the download or execution of external scripts or packages.
  • [Obfuscation] (SAFE): No encoded content, zero-width characters, or homoglyphs detected.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 12:45 AM