integrate-r3f-three-legacy
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes AI-generated 3D assets and code, creating an attack surface for indirect prompt injection from external sources like AI generators.
- Ingestion points: The agent ingests data through
read_fileandsearch_code. - Boundary markers: The skill explicitly instructs to treat AI generators as untrusted sources and uses component manifests for validation.
- Capability inventory: The agent has capabilities to modify files and execute build/test commands via
write_fileandrun_build. - Sanitization: The skill defines validation gates (ASI04, ASI05) to check polygon density, UV correctness, and material attribution for external assets.
Audit Metadata