ecommerce-common
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents patterns for ingesting and processing untrusted data from external sources, which constitutes a potential attack surface for indirect prompt injection.
- Ingestion points: The skill describes handling external data from sources such as payment provider webhook payloads (references/webhook-patterns.md), customer-provided checkout form fields (SKILL.md), and product data import files (references/product-data-models.md).
- Boundary markers: The instructions recommend several security boundaries, including HMAC-SHA256 signature verification for webhooks, CSRF token validation for all state-changing endpoints, and strict server-side validation for financial and custom checkout data.
- Capability inventory: The documented patterns involve powerful capabilities, including performing network requests to external APIs (payment gateways, shipping carriers), writing files to storage (shipping labels, digital product assets), and modifying database records (inventory levels, order statuses).
- Sanitization: The reference files (references/security-and-pci.md) emphasize critical sanitization and security practices, such as output encoding to prevent XSS, the use of parameterized queries to prevent SQL injection, and the sanitization of user-provided HTML content.
Audit Metadata