worldview-game-lure-hide-escape
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied project files, maps, and descriptions to generate gameplay code and mechanic contracts. This represents an attack surface for indirect prompt injection if malicious instructions are embedded in the game assets or project metadata.
- Ingestion points: Processes game projects, map layouts, sketches, and world packets supplied by the user as described in
SKILL.mdandREADME.md. - Boundary markers: The skill instructs the agent to inspect only authorized paths but does not specify explicit sanitization or delimiters for the ingested content to prevent the execution of embedded instructions.
- Capability inventory: Generates implementation code, writes to the file system, and may execute tests or browser automation.
- Sanitization: No explicit sanitization or validation of the ingested project data is mentioned in the provided instructions.
- [COMMAND_EXECUTION]: The skill may execute shell commands for testing purposes (e.g.,
npm testmentioned inexamples/last-footstep.md) and utilizes game engine runtimes (Unreal, Unity, Godot) or browser automation to verify encounters.
Audit Metadata