cs-escalation
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides comprehensive guidelines for managing escalated customer interactions, emphasizing empathy and structured response levels for different scenarios.
- [SAFE]: Privacy protection is integrated into the workflow, with explicit instructions to mask customer personal information (PII) before including it in analysis reports.
- [PROMPT_INJECTION]: The skill processes untrusted customer feedback (tickets, chat logs, SNS mentions), creating an indirect prompt injection surface. This risk is minimized by the skill's specific purpose and the recommendation for human-in-the-loop review for complex or emotionally charged grievances.
- Ingestion points: External customer feedback data sources defined in SKILL.md (VOC analysis) and references/feedback-analyzer.md (Data collection).
- Boundary markers: Absent; the skill does not specify delimiters for untrusted input data.
- Capability inventory: Generation of markdown reports and coordination with other vendor-specific post-processing skills.
- Sanitization: The skill includes directives for PII masking but does not specify technical sanitization to filter prompt injection attempts from ingested text.
Audit Metadata