test-game-exploits
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references security tools and technical documentation from well-known services and trusted organizations, including GitHub (mitmproxy, scanmem, AFL++), official Apple developer resources, and established game engine documentation (Unity, Unreal Engine, Godot).
- [PRIVILEGE_ESCALATION]: Instructions include using sudo to run memory auditing tools and adjusting Linux kernel security parameters like ptrace_scope. These actions are appropriately scoped to disposable test environments for the purpose of debugging the user's own software.
- [INDIRECT_PROMPT_INJECTION]: The skill provides defensive guidance for handling untrusted data. 1. Ingestion points: Client-side network messages and game save files. 2. Boundary markers: The instructions emphasize server-side authority to maintain a strict trust boundary. 3. Capability inventory: The skill utilizes tools like Cheat Engine, scanmem, and mitmproxy to audit software behavior. 4. Sanitization: The instructions recommend rigorous validation, including floating-point checks (math.isfinite) and strict schema validation for save loaders.
Audit Metadata