ad-campaign-analyzer

Pass

Audited by Gen Agent Trust Hub on Mar 28, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to process untrusted data from external sources (Google, Meta, and LinkedIn ad exports), which creates a surface for indirect prompt injection.
  • Ingestion points: Phase 0 and Phase 1 specify the ingestion of CSV files, pasted performance tables, and dashboard screenshots (SKILL.md).
  • Boundary markers: There are no explicit boundary markers or instructions to the agent to ignore embedded commands within the ingested data.
  • Capability inventory: The skill primarily uses reasoning but includes a file-writing instruction to save reports to 'clients//ads/'. It does not explicitly request high-risk tools like shell execution or network access.
  • Sanitization: The instructions do not define any sanitization or validation steps for the input data.
  • [NO_CODE]: The skill does not contain any executable scripts, binaries, or automated installation steps for third-party packages, relying entirely on the agent's internal reasoning capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 28, 2026, 11:42 AM