audio-setup-mixers

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill generates C# code snippets at runtime based on project state and executes them using the eval command in a live Unity Editor. This pattern is documented in SKILL.md (Step 0) and implemented in references/api.md (Assign groups section).\n
  • Evidence: The skill constructs a C# Dictionary and assignment logic from data discovered in the scene, which is then passed to unity command eval.\n- [INDIRECT_PROMPT_INJECTION]: The skill reads external, untrusted data from the Unity project (AudioClip and GameObject names) to guide its classification and code generation logic.\n
  • Ingestion points: Asset names are collected in references/api.md using UnityEngine.Object.FindObjectsByType.\n
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to treat asset names as potentially untrusted data or to escape them before insertion into shell commands or C# snippets.\n
  • Capability inventory: The skill can execute arbitrary C# code through the Unity CLI and has the ability to modify scene components and persist changes to the scene file.\n
  • Sanitization: The instructions do not define sanitization or validation routines for asset names prior to their use in code generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 05:47 PM
Security Audit — agent-trust-hub — audio-setup-mixers