The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user instructions to modify Unreal Engine assets. It lacks explicit instructions for the agent to sanitize or validate user-provided logic requirements before translating them into graph modifications, which could allow a malicious user to influence the generated Blueprint logic.
Evidence: The skill instructions prompt the agent to use the provided APIs to 'wire up Blueprint logic' or 'build an event graph' based on user requests, creating an attack surface for instructions embedded in the data processed by the agent.
[DYNAMIC_EXECUTION]: The skill generates and executes Python scripts using the unreal module to interact with the Unreal Engine editor environment. While this is the primary and intended purpose of the skill, the dynamic assembly of logic from user input and its subsequent execution via execute_python_code is a form of runtime code generation.
Evidence: The skill includes sample Python scripts (e.g., scripts/build_event_graph.txt) and instructions for building graphs using the build_graph API, which is then executed within the engine's Python context.

blueprint-graphs