Prediction Models

Prediction Models exposes ML model inference as tools for an AI Agent.

Purpose and use: Provides selected active ML model inference tools to an AI Agent. Use when the model needs to choose a configured prediction tool; this node itself is not a control-flow prediction action.

Configuration: Optional models string array, default empty, containing selected model identifiers.

Input/output and routing: Input input; outputs output and tool_out. Connect tool_out to AI Agent tools_in. Provider selection does not itself execute a prediction step.

Expressions: Model identifiers are static. Tool-call arguments are supplied by the AI Agent.

Errors, permissions, and effects: Missing/inactive selected models are skipped; tool invocation errors follow the AI Agent’s workflow policy. Availability and permissions depend on installation configuration. Inference returns recommendation data rather than updating a ticket; any write needs a separate action and its authorization.

Example configuration — fields validated against the 1.1.1 node schema; replace the identifiers with your own: {"models":[]}. Expected result: no model capability is exposed until you select one from the installed models.