customer-voice
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from external sources.
- Ingestion points: Call transcripts, meeting notes, and email threads retrieved via document and email search tools as described in
SKILL.md. - Boundary markers: The instructions do not define clear delimiters or "ignore embedded instructions" warnings for the content being extracted and processed.
- Capability inventory: The agent has access to
dispatchanddispatch_readonlytools for Salesforce operations, as well as email and document search capabilities. - Sanitization: There are no explicit sanitization, escaping, or validation steps defined for the customer quotes or email content before they are incorporated into the agent's context or final output.
Audit Metadata