wms-testing-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of comprehensive testing guidelines and code examples for validating WMS platforms such as SAP EWM, Blue Yonder, and Manhattan Associates. The content is educational and functional for quality engineering purposes.
- [SAFE]: No evidence of prompt injection, role-play bypasses, or attempts to override agent safety guidelines was found.
- [SAFE]: The skill does not contain hardcoded credentials, sensitive file path access, or unauthorized data exfiltration patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data sources including EDI documents (X12 856, 940, 945) and RF/barcode scanning inputs. This capability is the primary purpose of the skill. While this represents a potential surface for indirect prompt injection, it is a routine requirement for testing integration workflows.
- Ingestion points: edi.process940() and rfSession.scan() methods in SKILL.md process externally supplied data buffers.
- Boundary markers: No specific boundary markers are defined in the example scripts, but the logic is scoped to structural validation.
- Capability inventory: Capabilities are limited to simulated WMS API interactions and transaction validation through specialized agents.
- Sanitization: Standard validation logic for EDI segments and barcodes is demonstrated in the provided code snippets.
Audit Metadata