composites

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFE
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill includes a command-line script snippet using node -e in SKILL.md (Step 0b) to calculate 3D model bounding boxes by parsing GLB binary files. This is a localized utility designed for geometric validation during scene layout.
  • [COMMAND_EXECUTION]: The documentation references the use of npx sdk-commands build in SKILL.md to perform post-write validation of the scene. This is a standard and expected operation for users of the Decentraland SDK.
  • [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to ingest and process external project files, including scene.json and 3D model binaries (.glb), to compute scene bounds and placements. While this ingestion process presents an indirect prompt injection surface, the logic is constrained to technical geometric calculations within a developer workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 02:34 AM
Security Audit — agent-trust-hub — composites