The Agent Skills Directory

[SAFE]: The skill follows established security and development practices for the Godot engine. All identified patterns are dedicated to the primary purpose of console porting and certification.
[INDIRECT_PROMPT_INJECTION]: The skill defines ingestion points for user/game data during save and achievement operations, representing a minimal attack surface.
Ingestion points: execute_save in scripts/async_save_manager.gd and unlock_achievement in scripts/achievement_offline_queue.gd.
Boundary markers: Not required for these specific data-handling operations.
Capability inventory: File-write operations to the user:// local directory (standard for Godot save systems).
Sanitization: Employs JSON.stringify and store_var for serialization, preventing injection into the underlying data structure.
[COMMAND_EXECUTION]: Utilizes WorkerThreadPool.add_task in scripts/async_save_manager.gd and scripts/background_data_prefetcher.gd to perform background I/O. This is a standard Godot feature used here to prevent main-thread stutters during file saving and asset prefetching.
[DATA_EXPOSURE_AND_EXFILTRATION]: Local file access is restricted to the safe user:// sandbox for game state persistence. No network exfiltration or access to sensitive system paths (like .ssh or .aws) was found.

godot-platform-console