ad-creative

Pass

Audited by Gen Agent Trust Hub on Oct 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected. The skill provides marketing-focused instructions and utility scripts.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user-provided marketing briefs and campaign performance data.
  • Ingestion points: Data enters through manual brief creation in SKILL.md and through JSON files read by scripts/creative_fatigue_detector.py and scripts/creative_matrix_builder.py.
  • Boundary markers: The workflows use standard Markdown formatting to delimit inputs.
  • Capability inventory: The provided scripts are limited to string manipulation and mathematical calculations using the Python standard library. There are no network operations, file-write operations, or dynamic execution calls (exec/eval).
  • Sanitization: While the scripts do not explicitly sanitize input for prompt injection, the lack of privileged capabilities (no subprocess, no network) makes the attack surface non-exploitable.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 30, 2026, 05:06 PM