Run Prompt

Overview

ModelML allows you to automate research, data transformation, and logic chaining within workflows using the Run Prompt feature. By integrating versatile prompt steps into your Workflow Builder and referencing outputs throughout, you can chain AI-driven logic, customize grid population, automate variable updates, and enable multi-step analysis for streamlined operations at any scale.

For next steps after setting up the run prompt when you want to link the run prompt to the next workflow step:
The next trigger should be "Add row to grid" > Select the sheet name > Mode: Add multiple rows > Item to loop over: Structured output > Row Content: Structured Output
​

Run Prompt enables automated reasoning, structured data generation, and workflow orchestration by embedding AI-driven instructions directly into ModelML workflows, allowing teams to transform inputs, reference prior outputs, and execute multi-step analytical processes in a single operational environment. Workflow configuration begins by selecting the Run Prompt step inside Workflow Builder, which opens a prompt configuration panel where instructions, data sources, and output formats are defined to guide how the model interprets and processes inputs. Prompt authoring involves writing clear task instructions, selecting relevant source materials, and defining how responses should be structured so that outputs can be reliably reused across downstream workflow steps and operational logic. Dynamic input handling is achieved by inserting curly brackets to capture user-provided values and by using the @ symbol to reference grid data or previous Run Prompt outputs, enabling chained reasoning and contextual awareness across multiple workflow stages. Output configuration supports three primary formats—text for narrative responses, single structured outputs for standardized fields, and multiple structured outputs for repeated or tabular results— ensuring compatibility with different workflow and data population needs. Workflow integration occurs by linking Run Prompt outputs to subsequent steps, allowing automated updates to variables, grid population, logic branching, and further AI processing without manual intervention. Operational scalability is supported through the ability to chain prompts, reuse outputs, and standardize transformation logic, making it possible to automate research, data preparation, and analysis across large datasets and recurring processes. Process optimization emerges as teams refine prompts, structure outputs for reuse, and connect multiple Run Prompt steps to create end-to-end automated workflows that reduce manual effort while improving consistency and speed