ugc-campaign-management

Pass

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted User Generated Content (UGC) from Instagram and web forms, which serves as an attack surface for indirect prompt injection. \n
  • Ingestion points: External content is fetched from the Instagram Graph API (SKILL.md) and collected via direct upload widgets. \n
  • Boundary markers: There are no explicit markers or delimiters in the implementation to separate untrusted content from system instructions. \n
  • Capability inventory: The skill can send emails (scheduleEmail), post social media comments (postInstagramComment), and update database records (db.ugcSubmissions.upsert), which are potential targets for malicious input. \n
  • Sanitization: The inclusion of Perspective and Vision APIs provides brand safety moderation but does not specifically sanitize against prompt injection techniques.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 16, 2026, 11:50 AM