The Agent Skills Directory

[SAFE]: The skill content is purely instructional and focused on game development auditing. No malicious instructions, obfuscation, or safety bypasses were detected.
[NO_CODE]: No executable scripts, binaries, or shell commands are included in or referenced by the skill. The instructions are implemented as natural language guidance for the agent.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process project files and external checklists, which serves as an ingestion point for potentially untrusted data (SKILL.md). While it lacks explicit boundary markers or sanitization logic, its capabilities are limited to writing audit findings to specific local documentation paths (SKILL.md). This attack surface is inherent to the auditing purpose and is considered safe.

gsp-mechanics-systems-audit