roblox-building
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for ingesting and processing third-party assets from the Roblox Toolbox and generated content, which constitutes an indirect prompt injection surface.
- Ingestion points: External assets are loaded via IDs and URIs as described in
references/full.md. - Boundary markers: The instructions explicitly advise the user to inspect inserted models, check all nested scripts, and remove or disable unnecessary code before integration.
- Capability inventory: The skill utilizes
execute_luauto programmatically modify the Roblox environment. - Sanitization: It provides a validation script to verify property settings and structural integrity after asset placement.
- [SAFE]: The skill demonstrates security awareness by including official documentation links regarding third-party vulnerabilities and malicious script detection, encouraging users to verify asset provenance and code safety.
Audit Metadata