The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using npx to interact with the @minecraft/creator-tools package for rendering and viewing 3D models. This includes commands like rendermodel and view which operate on local project files.- [EXTERNAL_DOWNLOADS]: The skill triggers the download and execution of the @minecraft/creator-tools package from the NPM registry via npx. This package is a trusted resource associated with the official Minecraft development environment.- [DYNAMIC_EXECUTION]: The use of npx -y allows for the dynamic loading and execution of scripts at runtime, specifically targeting the official creator tools provided for Minecraft add-on development.- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from Minecraft project files, such as .entity.json and .geo.json. This represents an indirect prompt injection surface if these files were to contain malicious instructions.
Ingestion points: Reads entity configuration and geometry data from resource_packs/<pack>/entity/<id>.entity.json and <file>.geo.json.
Boundary markers: None identified in the prompt instructions for delimiting content from project files.
Capability inventory: The agent can execute shell commands via npx, write image files (writeImageFileFromPixelArt), and modify project structure through designModel.
Sanitization: No explicit sanitization or validation of the content read from the JSON project files is mentioned before the agent uses the data to perform actions.

design-model