The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data which could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: The scriptableobject_import_json skill accepts external data through the json string and jsonFilePath parameters.
Boundary markers: There are no delimiters or warnings defined to instruct the agent to ignore instructions embedded within the imported JSON data.
Capability inventory: The module possesses significant mutation capabilities, including creating (scriptableobject_create), modifying (scriptableobject_set, scriptableobject_set_serialized_property), and deleting (scriptableobject_delete) assets within the Unity project.
Sanitization: The instructions do not mention any validation, filtering, or escaping of the JSON content before it is applied to the ScriptableObject assets.

unity-scriptableobject