ecom-product-orchestrator
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources (Amazon, Temu, 1688) through its sub-skills, which presents a surface for indirect prompt injection if the external data contains instructions designed to manipulate the orchestrator's reporting or decision logic.
- Ingestion points: Data is ingested from the outputs of
amazon-movers-shakers,temu-competitor-search, andali1688-sourcingwithin the orchestrator's automated workflow. - Boundary markers: The output templates and JSON structures do not explicitly define boundary markers or 'ignore embedded instructions' prompts when interpolating external product data.
- Capability inventory: The skill's capabilities are limited to report generation, logic-based filtering (GO/PASS), and scoring; it does not directly execute shell commands or write to sensitive system files.
- Sanitization: There is no evidence of sanitization or validation of the strings returned from the external sub-skills before they are included in the final markdown report.
Audit Metadata