game-playbook
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-supplied game ideas into executable game code and assets, which presents a surface for indirect prompt injection.
- Ingestion points: Processes a "one-line idea" provided by the user in SKILL.md.
- Boundary markers: Absent; there are no explicit instructions to the agent to distinguish between the user's game idea and the agent's internal safety constraints.
- Capability inventory: The skill utilizes tools for file creation (
vg new), asset generation (vg generate), and deployment (vg deploy) as described in SKILL.md. - Sanitization: Absent; the skill provides code templates but does not suggest sanitization of user-provided strings before they are interpolated into game logic.
- [COMMAND_EXECUTION]: The skill instructs the agent to perform project management and deployment tasks via CLI tools.
- Evidence: Instructions include running
vg new <slug> --engine phaser,vg generatefor audio, andvg deploy ./distin SKILL.md. - Context: These are standard platform operations for project scaffolding and lifecycle management within the intended environment.
Audit Metadata