threejs-game-balance
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were identified. The skill is entirely instructional, defining a persona for game value calibration and workflow management.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions define an ingestion surface for external project data, which is a common pattern for design-focused agents.
- Ingestion points: Reads Work Items, project configurations, and Game Design Documents (GDD) as stated in
SKILL.md. - Boundary markers: None present in the text to delimit external data from instructions.
- Capability inventory: The skill is scoped to design and simulation tasks; no high-privilege tools, network access, or file-write permissions are requested.
- Sanitization: No sanitization of ingested content is mentioned.
Audit Metadata