Definition

In Agentria, a Function represents a branch extending from a node.

A function executes only when specific conditions are met, triggering the connected node and returning its result.

This allows a node to focus on a single task at a time, making the overall workflow more efficient.

Key Characteristics

• Conditional Execution: Functions are invoked only when predefined conditions are satisfied

• Extended Node Connections: Branch out from a node to form new execution paths

• Performance Optimization: Prevents nodes from performing unnecessary operations, improving efficiency

• Clear Decision-Making: Enables models to make more explicit choices through conditional branching

How It Can Be Used

For example, within a Python node, multiple functions can be defined and connected to separate LLM nodes.

Depending on user input or the state of the data, the appropriate function will be executed, sending results to its connected node.

These branched outputs can then either rejoin the original node flow or continue on to subsequent nodes, ensuring that the model performs tasks selectively based on conditions.