headless-intelligent-search
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows standard development patterns for search integration. All documented network requests target official VTEX infrastructure, specifically
sp.vtex.comandvtexcommercestable.com.br, which are legitimate vendor resources. No evidence of malicious code, hardcoded credentials, or obfuscation was found. - [INDIRECT_PROMPT_INJECTION]: The skill describes how to handle user-supplied search queries. While this is an entry point for untrusted data, the documented integration patterns are consistent with safe storefront search implementations.
- Ingestion points: User-provided
queryparameter inproductSearchandautocompletefunctions withinSKILL.md. - Boundary markers: The skill relies on standard URL parameterization and JSON structures.
- Capability inventory: Network requests via
fetchto vendor-controlled domains. - Sanitization: Not specifically required for this search-centric use case beyond standard API parameter handling.
Audit Metadata