The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes unity command eval to execute C# statements within the context of a live Unity Editor. This capability is used to interact with the Unity Editor API for tasks that cannot be performed through simple metadata edits, such as triggering asset re-imports via SaveAndReimport(). The skill is transparent about this mechanism and includes a SECURITY.md file acknowledging this high-privilege access.
[DYNAMIC_EXECUTION]: The skill generates C# code at runtime by populating templates with project-specific data (such as object names or asset paths). These snippets are then passed to the Editor's evaluation engine. This dynamic generation is restricted to the pre-defined 'recipes' provided in the skill's resource files.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the Unity project environment, such as asset names, paths, and component properties. This represents an attack surface where maliciously named assets could attempt to influence the agent's behavior. However, the skill primarily uses this data for read-only reporting and structured property modification, which significantly limits the potential for exploitation.

optimize-audio