cross-sell-upsell-detection
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes a workflow that ingests untrusted external data (such as customer usage metrics, feature requests, and external signals) to generate recommendations and outreach messages.
- Ingestion points: Data enters through
analyze_usage,analyze_engagement, andfetch_external_signalsfunctions described in theexpansion_detection_pipelineinSKILL.md. - Boundary markers: No explicit delimiters or instructions to ignore embedded instructions are present in the provided templates.
- Capability inventory: The skill is designed to generate outreach recommendations but does not contain capabilities for direct file system writes or network operations.
- Sanitization: The provided code snippets do not implement sanitization or validation for text-based signals (like "Explicit Mentions" or "Feature Requests") before they are interpolated into the message templates.
Audit Metadata