designing-minimal-game-rules
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a structured reasoning process for game design. It uses internal reference files (
breaker-roles.mdandfinal-output-template.md) to guide the AI through a multi-step evaluation process. - [PROMPT_INJECTION]: The skill contains strong imperative instructions such as 'Reject candidates', 'Do not apply', and 'must not silently repair'. These are task-specific constraints intended to enforce a particular game design philosophy (minimalism and discrete states) rather than attempts to bypass the underlying AI safety filters or system instructions.
- [DATA_EXPOSURE]: No sensitive file paths, environment variables, or hardcoded credentials were found. The skill only accesses its own local documentation files.
- [COMMAND_EXECUTION]: There are no shell commands, script executions, or subprocess calls defined in the workflow. The mention of 'subagents' refers to an LLM orchestration pattern for isolated reasoning, not local system execution.
- [INDIRECT_PROMPT_INJECTION]: The skill processes a user-provided 'seed' as input. While this represents an ingestion point for untrusted data, the skill lacks the capabilities (like network access or file writing) to be exploited via this vector. The output remains confined to the text conversation.
Audit Metadata