Copywriter

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOWNO_CODE
Full Analysis
  • [Prompt Injection] (SAFE): No attempts to override agent behavior or bypass safety constraints were found. The content consists of standard instructional workflows.
  • [Data Exposure & Exfiltration] (SAFE): The skill does not access sensitive local file paths or perform any network operations.
  • [Remote Code Execution] (SAFE): There are no scripts, binaries, or remote download patterns identified.
  • [Indirect Prompt Injection] (LOW): The skill processes user-provided input (product details, audience info). However, because it has no executable capabilities (no file-write, no network, no subprocesses), the surface for exploitation is minimal and limited to purely text-based reasoning.
  • [Obfuscation] (SAFE): No encoded strings, zero-width characters, or homoglyphs were detected.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 05:25 AM