customer-support

Pass

Audited by Gen Agent Trust Hub on Aug 9, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to process untrusted data from customer interactions (email, chat, social media) and has capabilities to perform write operations in external platforms, creating a surface for indirect prompt injection. |
  • Ingestion points: omnichannel customer communication including email, chat, social, and phone (SKILL.md). |
  • Boundary markers: No delimiters or instructions to ignore embedded commands were found in the skill body. |
  • Capability inventory: Integration with CRM systems (Salesforce, HubSpot), help desks (Zendesk, Freshdesk, Intercom), and custom API/webhook automation (SKILL.md). |
  • Sanitization: No specific validation or escaping procedures are defined for handling external inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 9, 2026, 08:24 PM
Security Audit — agent-trust-hub — customer-support