inttegro-resolve-customer-case
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill utilizes official vendor infrastructure for its toolset, connecting to the authenticated MCP server at https://mcp.inttegro.com/.
- [SAFE]: Strong privacy controls are enforced within the instructions, explicitly forbidding the exposure of masked contact information or payment credentials.
- [SAFE]: Destructive operations, such as payments, refunds, and order cancellations, require explicit merchant approval and are protected against accidental double-execution through the use of stable operation keys.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data like customer messages and order records, which represents a potential injection surface. However, the instructions mandate verification and consent boundaries that mitigate the risk of automated obedience to embedded malicious instructions.
Audit Metadata