build
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes store catalog data and section templates (fetched via the
lexsis-aiMCP server) alongside user prompts. This ingestion of external data constitutes a surface for indirect prompt injection, where malicious instructions embedded in templates or product descriptions could potentially influence the agent's behavior during the storefront build process.
Audit Metadata