---
title: Build your first workflow
display: 'Build your first *workflow*'
nav_title: Build a workflow
summary:
  - "Join a style prompt and a framing rule in a Prompt Builder, so each text is written once and reused."
  - "Wire the prompt and the character into an Image Generator on GPT Image 2.5 Sunburst, so it redraws her instead of inventing someone new."
  - "Test one step with Run This Node before you run the whole workflow; the price shows on the button."
description: Add two Text nodes, join them in a Prompt Builder, send the prompt and the character to an Image Generator, and test that one step to make the first anatomy sheet.
section: workflows
order: 40
type: tutorial
level: Build
roles: [all]
time: { read: 9, do: 15, wait: 1 }
needs: ["The Anatomy reference sheets workflow from the previous lesson", "One character image in your library"]
last_verified: 2026-11-11
models_checked: ["GPT Image 2.5 Sunburst"]
hero: workflows/build-your-first-workflow/step-07-first-sheet-on-the-node.webp
related: [workflow-editor-tour, nodes-reference, prompts-and-llm-nodes, debug-a-workflow, publish-a-workflow-as-an-app, platform-workflows-and-templates]
thumb: workflows/build-your-first-workflow/thumb.webp
video: tutorials/build-your-first-workflow.mp4
video_caption: "Build your first workflow"
chapters:
  - { at: 0, title: "The anatomy sheet to make", image: "/media/images/workflows/build-your-first-workflow/ch-01-the-anatomy-sheet-to-make.webp" }
  - { at: 23, title: "Open the workflow", image: "/media/images/workflows/build-your-first-workflow/ch-02-open-the-workflow.webp" }
  - { at: 44.5, title: "Add a Text node", image: "/media/images/workflows/build-your-first-workflow/ch-03-add-a-text-node.webp" }
  - { at: 72.5, title: "Add the style text", image: "/media/images/workflows/build-your-first-workflow/ch-04-add-the-style-text.webp" }
  - { at: 92.5, title: "Add a Prompt Builder", image: "/media/images/workflows/build-your-first-workflow/ch-05-add-a-prompt-builder.webp" }
  - { at: 129, title: "Add an Image Generator", image: "/media/images/workflows/build-your-first-workflow/ch-06-add-an-image-generator.webp" }
  - { at: 152.5, title: "Choose the model", image: "/media/images/workflows/build-your-first-workflow/ch-07-choose-the-model.webp" }
  - { at: 165, title: "Run this node", image: "/media/images/workflows/build-your-first-workflow/ch-08-run-this-node.webp" }
  - { at: 177, title: "Check the sheet", image: "/media/images/workflows/build-your-first-workflow/ch-09-check-the-sheet.webp" }
  - { at: 199.5, title: "Recap", image: "/media/images/workflows/build-your-first-workflow/ch-10-recap.webp" }
steps:
  - title: "Open the workflow"
    text: "Open **Workflows** › **Workflows**, **Project** tab, and open **Anatomy reference sheets**. Click **Switch to Edit Mode**."
  - title: "Make room"
    text: "Drag an empty part of the canvas to slide the character to the left, and scroll to zoom."
  - title: "Add the framing rule"
    text: "Click **+** › **Input** › **Text** (Shift + T), name it `1. Full Body + 9:16` and paste the framing rule. It asks for a tall 9:16, head-to-toe frame."
    prompt: |-
      Output the image in a 9:16 vertical portrait aspect ratio.

      If the source image is cropped (bust, half-body, or cut off at any edge), extend it into a COMPLETE FULL-BODY figure: reconstruct the missing torso, hips, legs and feet in a natural continuation of the pose, and frame the whole figure head-to-toe, centered in the 9:16 vertical frame with a small margin of empty space above the head and below the feet. Keep the same character design, proportions, style and camera direction. If the source already shows the full body, keep the pose and framing, just fit it into the 9:16 vertical frame.
  - title: "Add the style prompt"
    text: "Add a second **Text** node above it, name it `2. Style C - Anatomy` and paste the style prompt: a pencil sketch with the body drawn in blue shapes."
    prompt: |-
      Redraw this character illustration as a traditional graphite pencil sketch on textured off-white paper.

      Keep the exact same pose, proportions, silhouette, outfit, camera angle and framing as the source.

      PRESERVE EVERY ELEMENT OF THE CHARACTER: capes, cloaks, hoods, belts, straps, armor, gloves, boots, bags, weapons, wings, tails, jewelry and hair must all be drawn, with the same shape, length and placement as in the reference. Do not remove or simplify any part of the design.

      Hand-drawn look: confident construction lines, light underdrawing still visible, expressive contour linework, no cross-hatching. Monochrome graphite only — no color, no digital gradients, no background scenery, just clean paper.

      Overlay a drawing of the body anatomy drawn in basic geometry (spheres and cylinders etc.) in clean blue pencil line, revealing the body shape hidden by clothes and equipment.

      Style: professional character design study sketch from an artist's sketchbook, figure drawing reference, no text, no watermark.
  - title: "Add a Prompt Builder"
    text: "Under **Utilities**, add a **Prompt Builder** and name it `3. Final Prompt C`."
  - title: "Wire in the two texts"
    text: "Drag from the style's blue output to the Prompt Builder's text port, then from the framing rule. Each text appears as a chip."
  - title: "Build the prompt"
    text: "In the template box, click the style chip, press Enter twice, then click the framing chip. A chip not placed in the template isn't sent."
  - title: "Add an Image Generator"
    text: "Under **Generators**, add an **Image Generator** to the right and name it `4. Output C - Anatomy`. It has a blue **Prompt** port and a green **Reference Images** port."
  - title: "Wire the prompt and the character"
    text: "Wire the Prompt Builder into **Prompt** and the character's green **IMAGE** port into **Reference Images**, so the model redraws her."
  - title: "Choose the model"
    text: "Click the model name on the node, search `Sunburst` in **Choose a model**, select **GPT Image 2.5 Sunburst** and click **Use this model**."
  - title: "Run this node"
    text: "Leave **Image Count** at 1 and **Dimensions**, **Quality** and **Background** on Auto. Click **Run This Node** (Cmd + Enter) to run only this step."
  - title: "Check the sheet"
    text: "Open it full size. Accept the same pose, every piece of gear, graphite on off-white paper with blue body shapes, head to toe; reject missing gear, a new character, extra color or text."
  - title: "Fix it, then close"
    text: "Fix the prompt while a test costs one image. The workflow saves as you work, so close it with the **X** at the top left when you're done."
updated: 2026-11-11
canonical: https://learn.scenario.com/workflows/build-your-first-workflow
---

Last updated: 2026-11-11

# Build your first workflow

> Add two Text nodes, join them in a Prompt Builder, send the prompt and the character to an Image Generator, and test that one step to make the first anatomy sheet.

In short:

- Join a style prompt and a framing rule in a Prompt Builder, so each text is written once and reused.
- Wire the prompt and the character into an Image Generator on GPT Image 2.5 Sunburst, so it redraws her instead of inventing someone new.
- Test one step with Run This Node before you run the whole workflow; the price shows on the button.


This lesson makes the first of the four sheets in **Anatomy reference sheets**: a pencil sketch of the character with the anatomy drawn on top in blue. You add two texts, join them into one prompt with a **Prompt Builder**, send that prompt and the character to an **Image Generator** on GPT Image 2.5 Sunburst, and test that one step on its own.

:::note As of November 2026
The models, settings and screens in this article are the ones Scenario offered in November 2026. New models and updates arrive often, so check the app for newer options before you choose.
:::

:::steps Quick version
1. Open **Workflows** › **Workflows**, and on the **Project** tab open **Anatomy reference sheets**. Click **Switch to Edit Mode**.
2. Add two **Text** nodes: `1. Full Body + 9:16` with the framing rule, and `2. Style C - Anatomy` with the style prompt.
3. Add a **Prompt Builder** named `3. Final Prompt C`, wire both texts into it, and build the prompt: style chip, blank line, framing chip.
4. Add an **Image Generator** named `4. Output C - Anatomy`. Wire the prompt into **Prompt** and the character into **Reference Images**, and choose **GPT Image 2.5 Sunburst**.
5. Click **Run This Node** and check the sheet. The workflow saves as you work (Cmd+S, or Ctrl+S on Windows, saves right away); close it with the **X** when you're done.
:::

## Before you start

- **What you need:** the **Anatomy reference sheets** workflow from [What workflows are](https://learn.scenario.com/workflows/what-are-workflows.md), with its **1. Character Illustration** node holding your character.
- **What it costs:** only the Image Generator spends anything in this lesson, one image. The price shows on **Run This Node** before you click.
- **Why two texts and not one prompt:** the framing rule is the same for all four sheets, the style is not. Keeping them apart means the rule is written once and every look reuses it. The next lesson builds on this.

### The nodes in this lesson

| Node | Kind | What it holds or does |
|---|---|---|
| **1. Character Illustration** | Image input, from the previous lesson | The character |
| **1. Full Body + 9:16** | Text input | The framing rule, shared by all four looks |
| **2. Style C - Anatomy** | Text input | The style prompt for this look |
| **3. Final Prompt C** | Prompt Builder | Joins the style and the rule into one prompt |
| **4. Output C - Anatomy** | Image Generator | Redraws the character with GPT Image 2.5 Sunburst |

## Step 1: Open the workflow

Open **Workflows** › **Workflows**, and on the **Project** tab, open **Anatomy reference sheets**. A workflow in your project opens in run-only mode. Click **Switch to Edit Mode**.

![The run-only banner: You're in run-only mode, with the Switch to Edit Mode button](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/e617ac5481a4550fffcd088cdb92d9de79cbfb9c66c131a29e695f7d941a8a2d.webp)

## Step 2: Make room

Drag an empty part of the canvas to slide the character over to the left. That leaves room on the right for the next nodes. Scroll to zoom if the nodes look too small or too big.

## Step 3: Add the framing rule

Click **+**, and under **Input**, choose **Text** (or press Shift + T). Drag it into place to the right of the character.

Double-click its title and name it:

```text
1. Full Body + 9:16
```

The number keeps the graph in reading order. Click inside the node and paste the framing rule. It asks for a tall 9:16 frame with the whole body, head to toe:

```prompt
Output the image in a 9:16 vertical portrait aspect ratio.

If the source image is cropped (bust, half-body, or cut off at any edge), extend it into a COMPLETE FULL-BODY figure: reconstruct the missing torso, hips, legs and feet in a natural continuation of the pose, and frame the whole figure head-to-toe, centered in the 9:16 vertical frame with a small margin of empty space above the head and below the feet. Keep the same character design, proportions, style and camera direction. If the source already shows the full body, keep the pose and framing, just fit it into the 9:16 vertical frame.
```

The node shows a character count under the text: 598 characters for this rule.

## Step 4: Add the style prompt

Add a second **Text** node above the first one and name it:

```text
2. Style C - Anatomy
```

Paste the style prompt: a pencil sketch of the same pose, with the body drawn in simple blue shapes over the clothes. The next lesson, [Prompts that repeat](https://learn.scenario.com/workflows/prompts-and-llm-nodes.md), explains how it's written.

```prompt
Redraw this character illustration as a traditional graphite pencil sketch on textured off-white paper.

Keep the exact same pose, proportions, silhouette, outfit, camera angle and framing as the source.

PRESERVE EVERY ELEMENT OF THE CHARACTER: capes, cloaks, hoods, belts, straps, armor, gloves, boots, bags, weapons, wings, tails, jewelry and hair must all be drawn, with the same shape, length and placement as in the reference. Do not remove or simplify any part of the design.

Hand-drawn look: confident construction lines, light underdrawing still visible, expressive contour linework, no cross-hatching. Monochrome graphite only — no color, no digital gradients, no background scenery, just clean paper.

Overlay a drawing of the body anatomy drawn in basic geometry (spheres and cylinders etc.) in clean blue pencil line, revealing the body shape hidden by clothes and equipment.

Style: professional character design study sketch from an artist's sketchbook, figure drawing reference, no text, no watermark.
```

![The two Text nodes: 2. Style C - Anatomy at 1018 characters above 1. Full Body + 9:16 at 598 characters](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/0c448963161dfb2d0920248954d7f2cc0494fad90e3a7b8e1e564e2ad62d0179.webp)

## Step 5: Add a Prompt Builder

A **Prompt Builder** joins several texts into one prompt through a template. Under **Utilities**, add a **Prompt Builder**. Click its title once and name it:

```text
3. Final Prompt C
```

## Step 6: Wire in the two texts

Drag from the style's blue output dot to the Prompt Builder's text port on its left edge. Then do the same with the framing rule. Each text shows up as a chip at the top of the node: **2. STYLE C - ANATOMY** and **1. FULL BODY + 9:16**.

## Step 7: Build the prompt

Click in the template box. Click the style chip, press Enter twice for a blank line, then click the framing chip. Underneath, the node shows the whole prompt it will send: the style text, a blank line, then the framing rule.

![3. Final Prompt C with its two chips and the style chip placed in the template, wired from both Text nodes](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/3f65bad3eb2bd1ceccbc57584561dcac16875ad0b4d5e1c8931e7898565d45f0.webp)

You can also type your own words around the chips: they stay fixed on every run. A chip that isn't placed in the template is not sent.

## Step 8: Add an Image Generator

Under **Generators**, add an **Image Generator**. Drag it to the right of the Prompt Builder and name it:

```text
4. Output C - Anatomy
```

Its ports, on the left edge:

| Port | Color | What goes in |
|---|---|---|
| **Prompt** | Blue | The text the model follows |
| **Reference Images** | Green | Images the model works from. An Image node with several images passes up to 10 |

On the right edge, its output carries the generated image.

## Step 9: Wire the prompt and the character

Wire the Prompt Builder's output into **Prompt**. Then wire the character's green **IMAGE** port into **Reference Images**, so the model redraws her instead of inventing someone new.

![The settings panel of 4. Output C - Anatomy next to the graph: the prompt wired into Prompt and the character into Reference Images](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/f3090aad6f11e15fd9d1b9b6d7c73daa2b2b9692a872dc575354e115c837300b.webp)

Click the node to open its settings in the panel on the left. **Prompt** now shows the joined text that comes from the Prompt Builder.

## Step 10: Choose the model

Click the model name at the bottom of the node. In **Choose a model**, search for `Sunburst`, select **GPT Image 2.5 Sunburst**, and click **Use this model**. It redraws an image and keeps the pose.

![Choose a model searched for Sunburst: GPT Image 2.5 Sunburst, selected, and GPT Image 2.6 Sunburst](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/0afee6e277498a0ef6a51a82f011a43fe7b61c9cb50cd14a36309c8be483a543.webp)

The chooser lists every image model on Scenario, including the models your team trained, with filters by input, provider and capability. Each row shows the provider, a rank and a typical time. Any model fits in the node, and swapping it later doesn't change the rest of the graph.

## Step 11: Run this node

The panel on the left now shows the model, the prompt and the reference. Leave the settings at their defaults:

| Setting | What it does | In this lesson |
|---|---|---|
| **Image Count** | How many images one run makes, from 1 to 10 | 1 |
| **Dimensions** | The output size | Auto |
| **Quality** | The model's quality level | Auto |
| **Background** | The model's background option | Auto |

![The Image Generator panel: Model GPT Image 2.5 Sunburst, the joined Prompt, Reference Images, and Settings](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/2ef13fb86b97b2d9353a40d3445a7765a3518bb36fd67899741bd040dc7bf3ab.webp)

Click **Run This Node** at the bottom of the panel (Cmd + Enter). It runs only this step, not the whole workflow. The node shows that it's running, then the sheet appears on it.

| Run button | What runs |
|---|---|
| **Run This Node** | Only this node |
| **Run From Here** | This node and everything after it |
| **Run** (top right) | The whole workflow, each node as soon as its inputs are ready |

## Step 12: Check the sheet

The first sheet lands on the node. Open it full size with the enlarge button on the image.

![The first sheet on the 4. Output C - Anatomy node, after Run This Node](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/1ce92d15ea4ce54b0839d089189e63f8d397c321340f3570f7c73e52151dbdec.webp)

![The first sheet full size: the ranger as a graphite pencil sketch, with blue construction lines for the body under the clothes](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/09c4f65066779008692f3d344eddce7cf926932ae07b3768503504cfe17db510.webp)

Same pose, every piece of her gear, and blue shapes for the body under the clothes.

:::check
**Accept when:** the pose, proportions and camera angle match the character; every piece of gear is there (hood, cape, shoulder armor, belts, sword, shield, boots); the drawing is graphite on off-white paper; blue lines show the body's shapes; the frame is tall and shows her head to toe.
**Reject when:** a strap, the cape or a weapon is missing; she's a different person or in a different pose; there's color besides the blue lines; the feet or head are cut off; text or a watermark appears.
:::

## Step 13: Fix it, then close

If something's missing, fix the prompt now, while a test costs one image, and run the node again. The workflow saves as you work, and Cmd+S (Ctrl+S on Windows) saves right away if you want to be sure. When you're done, close it with the **X** at the top left.

![The X and the workflow title at the top left of the editor](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/d7ed4695eafa0f73f63b8d4576f607f94e1fa22543861fcfe47713226c661fd4.webp)

## How the Prompt Builder works

- **One output, many inputs.** Connect as many text inputs as you need. Each appears as a chip.
- **The template decides the order.** Place chips in any order, with line breaks and your own fixed words between them.
- **It takes a model's text too.** A chip can come from an **LLM Node**'s result as well as from a **Text** node.
- **Keep it short enough.** Each input holds about 1,000 to 2,000 characters, and the joined prompt about 5,000, depending on the model it goes to.

## Common problems

| What you see | What to do |
|---|---|
| **You can't add or move nodes** | The workflow is in run-only mode. Click **Switch to Edit Mode** |
| **A wire won't connect** | The port colors differ. The Prompt Builder's blue output goes into the blue **Prompt**, the green character into green **Reference Images** |
| **The prompt underneath shows only the style** | The framing chip isn't placed in the template. Click in the box, then click the chip |
| **The sheet shows a new character** | The character isn't wired into **Reference Images**. Check the green wire |
| **Gear is missing from the sheet** | Run the node again, or make the PRESERVE list in the style prompt name the missing part |
| **She's cropped at the knees** | Check that the framing chip is in the template and that **1. Full Body + 9:16** still holds its text |
| **The run costs more than you expect** | Check **Image Count** is 1 before you click **Run This Node** |

## What's next

One look is done. In [Prompts that repeat](https://learn.scenario.com/workflows/prompts-and-llm-nodes.md), you read how the style prompt is written and add the prompts for the other three sheets. For every node's ports and settings, see the [Nodes reference](https://learn.scenario.com/workflows/nodes-reference.md); when a step fails, see [Debug a workflow](https://learn.scenario.com/workflows/debug-a-workflow.md).

Source: https://learn.scenario.com/workflows/build-your-first-workflow
