personalization-engine
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill serves as an implementation guide for e-commerce personalization and does not contain malicious code, obfuscation, or unauthorized data access patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill describes the implementation of a system that ingests untrusted user data such as browsing history and session data.
- Ingestion points: User session identifiers (sessionProductIds) and order history databases (order_items) referenced in SKILL.md.
- Boundary markers: The implementation logic focuses on mathematical vector similarity and database aggregations, which naturally separates data from execution instructions.
- Capability inventory: The described system involves database read/write access and Redis caching as outlined in SKILL.md and the evaluation tasks.
- Sanitization: Standard database querying practices are implied, but the templates do not explicitly show input sanitization for session-derived data.
Audit Metadata