audio-design
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No prompt injection patterns, overrides, or jailbreak attempts were detected in the instructions.
- [DATA_EXFILTRATION]: No sensitive file access or unauthorized network operations were identified.
- [OBFUSCATION]: No hidden content, encoded strings, or homoglyph attacks were found in the skill files.
- [REMOTE_CODE_EXECUTION]: No remote scripts are downloaded or executed; all code samples are local GDScript patterns for audio management.
- [INDIRECT_PROMPT_INJECTION]: The skill does not ingest untrusted external data that could influence agent behavior. Logic is confined to internal game state variables.
Audit Metadata