product-support-specialist
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes customer tickets and user feedback, which represent untrusted external data sources.
- Ingestion points: External content enters the agent's context through customer inquiries and feature requests as described in
references/triage_routing.mdandreferences/feedback_feature_requests.md. - Boundary markers: The instructions provide structured templates for feedback capture and ticket triage, which serve as organizational delimiters for incoming data.
- Capability inventory: The skill instructions and reference files do not include any commands for file system writes, network operations, or shell execution.
- Sanitization: No specific methods for sanitizing or filtering input from customer tickets are mentioned in the procedural guides.
Audit Metadata