The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides tools such as unity_create_script and unity_update_script that allow an AI assistant to write or modify C# code. This code is automatically compiled and executed within the Unity Editor environment when the project is in play mode or during standard editor operations.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted data from the local project environment which could influence AI behavior.
Ingestion points: unity_read_script, unity_get_scene_hierarchy, unity_get_console_logs, and unity_list_assets ingest project-specific strings and file contents into the AI context.
Boundary markers: The instructions do not define delimiters or specific warnings to ignore instructions embedded in the project data.
Capability inventory: Powerful tools like unity_update_script, unity_build, unity_play, and unity_hub_install_editor provide significant impact if the AI is successfully subverted.
Sanitization: There is no mention of filtering or sanitizing C# comments or asset names before they are processed by the AI.
[EXTERNAL_DOWNLOADS]: The installation process involves cloning the server repository from github.com/AnkleBreaker-Studio/unity-mcp-server.git and installing a Unity plugin from github.com/AnkleBreaker-Studio/unity-mcp-plugin.git. These resources are provided by the skill author.
[COMMAND_EXECUTION]: The skill interacts with the host system by executing commands via the Unity Hub CLI (unity_hub_install_editor) and running build processes (unity_build) that generate executable files.

unity-mcp-server