Skip to main content
The Playground is where you run AI applications built in Giselle. It provides an intuitive interface for executing apps, entering inputs, and viewing results.

​
Accessing the Playground

Navigate to Playground from the sidebar under “Stage - Run Apps”.

​
Interface Overview

​
App Selection

The Playground displays available apps organized by source:
  • Sample app: Pre-built example applications provided by Giselle
  • Your team app: Apps created by members of your team
Each app card shows:
  • App name and description
  • LLM providers used (OpenAI, Anthropic, Google, etc.)
  • Creator information
  • Source badge
Use the search bar to filter apps by name when you have many apps to choose from.

​
Input Area

Once you select an app, you can provide inputs: Text Input Enter your task description or prompt in the text area. This is the main input that tells the app what you want it to do. File Attachments Click the paperclip icon to attach files. Supported file types:
  • PDF documents
  • Images (PNG, JPG, JPEG, GIF, BMP, WEBP, SVG, TIFF, HEIC, HEIF)
You can also drag and drop files directly onto the input area.

​
Running an App

To run an app:
  1. Select an app from the available options
  2. Enter your task description in the text input
  3. Optionally attach relevant files
  4. Press Enter or click the send button to start
The Playground will show an overlay indicating the app is running, then redirect you to the task results page.

​
Viewing Results

After running an app, you’ll be redirected to the task page where you can view:
  • The generated output
  • Execution details
  • Any intermediate steps or processing information

​
Task History

Access your previous runs from Task History in the sidebar. This allows you to:
  • Review past task results
  • Compare different runs
  • Track your app usage over time

​
Tips for Using the Playground

  • Start with sample apps to understand how different workflows operate
  • Use descriptive prompts to get better results from your apps
  • Attach relevant files when your task requires document analysis or image processing
  • Check the app description to understand what inputs each app expects