The Agent Skills Directory

[SAFE]: The skill provides deep technical expertise and auditing for Godot 4.7+ projects. It explicitly warns against security anti-patterns like untrusted object decoding.
[COMMAND_EXECUTION]: The skill utilizes a fleet of Python and GDScript diagnostic tools (e.g., audit_signals.py, audit_memory_fragmentation.gd) to perform local code analysis and memory profiling. These tools operate strictly on the local project directory and lack network access.
[INDIRECT_PROMPT_INJECTION]: The skill acts as an auditor for external project files (GDScript, scene files, and configurations). This behavior constitutes an attack surface for indirect prompt injection if project files contain adversarial instructions. However, the logic is limited to pattern-matching and regex-based auditing, reducing the threat to low.

godot-auditor