Condition Agent Node

Provides AI-driven dynamic branching based on natural language instructions for situational decision making.

Functionality

This node uses a Large Language Model (LLM) strictly routed explicitly for text classification. It analyzes provided input data against a set of user-defined "Scenarios" — guided by high-level natural language instructions. The LLM then deterministically votes on which scenario semantically best fits the current contextual input. This node is critical for dynamic tasks like initial user-intent recognition (e.g., routing a user to "Sales", "Support", or "Billing" based on their raw chat input).

Configuration Parameters

• Model: Specifies the AI model tasked purely with performing the NLP scenario classification.
• Instructions: Define the overall algorithmic goal or task for the LLM in natural language — e.g., "Determine if the user's request is regarding an urgent production outage."
• Input: Specify the data utilizing variables ({{ chat_history }}) that the NLP engine will evaluate.
• Scenarios: Configure an array defining the distinct possible routing outcomes or branches. Each scenario is named using natural language — e.g., "Urgent Incident", "General Inquiry" — and each automatically spans a unique physical output anchor node on the canvas block.

Inputs & Outputs

• Inputs: Natively consumes text parameters mappings.
• Outputs: Physical output routes matching exactly the designated scenarios.