Custom Pop Ups
Custom Pop Ups are configurable instructional dialogs that appear at key moments during the VO to Images pipeline. They let workspace admins provide step-by-step guidance to team members — and are purpose-built for AI browser automation tools like Claude Cowork and OpenClaw AI.
Custom Pop Ups are designed to work with AI browser agents that can read on-screen instructions and follow them automatically — turning your VTI workflow into a hands-free production pipeline.
Why Custom Pop Ups?
When producing content at scale, teams often repeat the same workflow steps across dozens of projects. Custom Pop Ups let you:
- Standardize processes — Write instructions once, and every team member (or AI agent) follows them consistently
- Enable AI automation — Tools like Claude Cowork and OpenClaw AI can detect and read popup content via the
data-testid="custom-popup"attribute, then execute the instructions automatically - Reduce errors — Remind users of critical settings or steps that are easy to forget (e.g., toggling Smart Split, adjusting pacing)
Trigger Types
There are two types of popup triggers:
Step Navigation Popups
These appear when a user navigates to a specific step in the VTI pipeline:
| Trigger | When it fires |
|---|---|
| Entering Step 1 — Audio | User navigates to the Audio step |
| Entering Step 2 — Characters | User navigates to the Characters step |
| Entering Step 3 — Models | User navigates to the Models step |
| Entering Step 4 — Segments | User navigates to the Segments step |
| Entering Step 5 — Storyboard | User navigates to the Storyboard step |
| Entering Step 6 — Images | User navigates to the Images step |
| Entering Step 7 — Videos | User navigates to the Videos step |
| Entering Step 8 — Export | User navigates to the Export step |
Event Completion Popups
These appear when an asynchronous operation finishes:
| Trigger | When it fires |
|---|---|
| Segments Analyzed | Smart Detect finishes analyzing the transcript into segments |
| Characters Detected | Per-sentence character detection completes |
| Prompts Generated | Batch prompt generation finishes for all segments |
| Images Generated | All images in the current batch finish generating |
| Videos Generated | All videos in the current batch finish generating |
Setting Up Custom Pop Ups
Open VTI Settings
Navigate to the VO to Images page and click the Settings (gear) icon. You'll see tabs for Styles, Presets, and Custom Pop Ups.
The Custom Pop Ups tab is only visible to users with the manage_custom_popups permission. See Roles & Permissions for how to grant this permission.
Select a Trigger
In the Custom Pop Ups tab, you'll see all available triggers organized into two sections: Step Navigation Popups and Event Completion Popups. Click on a trigger to expand it.
Configure the Popup
For each trigger, you can configure:
- Enable/Disable toggle — Turn the popup on or off without deleting the content
- Popup Title — A short heading shown at the top of the dialog (e.g., "Before you start segmenting...")
- Popup Content — Rich text body with formatting support (bold, links, numbered lists, etc.)
Save
Click Save to store the popup configuration. The popup will now appear for all workspace members when the trigger fires.
Granting Permission
By default, only the workspace Director (owner) can manage Custom Pop Ups. To let other team members configure popups:
- Go to Settings > Roles & Permissions
- Select the role you want to grant access to (e.g., Admin, Editor)
- Enable the Manage Custom Pop Ups permission
- Click Save
Users with this permission will see the Custom Pop Ups tab in VTI Settings. Users without the permission will still see the popups during the workflow — they just can't edit them.
Use Case: AI Browser Automation
The primary use case for Custom Pop Ups is enabling AI browser automation tools to follow structured instructions during the VTI pipeline.
Example Workflow
Say you want an AI agent to handle the segmentation step. You'd configure a popup for Entering Step 4 — Segments with content like:
Instructions for AI:
- Click "Smart Detect" with Smart Split toggled ON
- Wait for analysis to complete
- Review the segment boundaries — merge any segments shorter than 2 seconds
- For the first 2 minutes of content, set pacing to 4-5 seconds per image
- Click "Next" to proceed to Storyboard
When an AI browser agent (Claude Cowork, OpenClaw AI) navigates to the Segments step, it detects the popup, reads the instructions, and executes them step-by-step.
Tips for Writing AI-Friendly Instructions
- Be specific — Use exact button names and UI element descriptions
- Number your steps — AI agents follow numbered lists more reliably than paragraphs
- Include wait conditions — Tell the agent what to wait for (e.g., "Wait for the loading spinner to disappear")
- One action per step — Keep each step to a single clear action
- Use the event popups — For operations that take time (image generation, video generation), configure an event completion popup with the next set of instructions