unity-asset-integration
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute several shell-based commands to automate the asset pipeline. This includes running local Python scripts (
tools/3d/deploy_accepted_model.pyandtools/3d/animation_quality/creatures/unity_install.py) and invoking the Unity editor executable in batch mode with specific command-line arguments and C# methods (-executeMethod).\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local project manifests and configuration files to determine installation parameters, paths, and metadata. This introduces a surface where potentially untrusted data from the repository could influence the behavior of the automation scripts.\n - Ingestion points: The skill reads
ArtSource/3D/Roster/subjects.json,ArtSource/3D/Meshy/manifest.json,animation-manifest.json, andProjectSettings/ProjectVersion.txt.\n - Boundary markers: No explicit markers or delimiters are defined in the instructions to prevent the agent from obeying instructions that might be embedded in these data files.\n
- Capability inventory: The agent has the capability to execute Python scripts, run shell commands, and execute arbitrary C# code within the Unity editor context via
-executeMethod.\n - Sanitization: There is no evidence of sanitization or validation logic described in the instructions to filter content from the ingested files before processing.
Audit Metadata