unity-implement

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from task files that could potentially contain malicious instructions intended to manipulate the agent's behavior.
  • Ingestion points: The skill instructs the agent to read task descriptions from files matching the pattern docs/<feature-slug>/tasks/<NN>-<slug>.md in the SKILL.md file.
  • Boundary markers: There are no explicit boundary markers or instructions to treat the content of the task files strictly as data rather than instructions.
  • Capability inventory: The skill possesses significant capabilities through the /unity-cli tool, including the ability to open the Unity Editor (unity open), recompile C# code, modify project assets (scenes, prefabs, .meta files), and list pipeline configurations in SKILL.md.
  • Sanitization: The instructions do not define any sanitization or validation steps for the content ingested from the external documentation files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 12:12 AM
Security Audit — agent-trust-hub — unity-implement