The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided natural language descriptions to generate GDScript files. This creates an attack surface where malicious input could influence the generated code to perform unintended or harmful actions when executed within the Godot environment.
Ingestion points: Natural language input describing game mechanics in SKILL.md (e.g., "Role Control Development", "AI Behavior Development").
Boundary markers: Absent. There are no instructions for the agent to sanitize or ignore potentially malicious sub-instructions within user requests.
Capability inventory: High capability tools are referenced in SKILL.md, including create_script, edit_script, attach_script, and play_scene.
Sanitization: Absent.
[DYNAMIC_EXECUTION]: The primary purpose of the skill involves generating and modifying executable script files via create_script and edit_script (SKILL.md). While this is the intended functionality for a developer tool, it represents a high-risk capability if the generation logic is influenced by malicious input.
[COMMAND_EXECUTION]: The skill utilizes MCP tools like play_scene to interact with and execute code within the active game engine environment, which can lead to arbitrary code execution if generated scripts are malicious.

godot-gameplay-developer