Use JavaScript Code
JS Code steps let you run custom JavaScript inside your workflow — useful for data manipulation, calculations, conditional logic, or calling external APIs.
JS Code runs as an action step. It executes after your trigger and can use data from any previous step as input.
When to Use JS Code
Perform calculations or transform data before passing it to the next step
Handle logic that built-in actions can't cover
Call external APIs and process the response
Filter, format, or restructure input data based on conditions
How to Add a JS Code Step
1. Add a JS Code action
In the workflow editor, click + Add Step. In the panel that opens, scroll to Built-In Tools and select JS Code.

2. Describe your requirement using Ask AI
Once the JS Code step opens, you'll see a text area with the placeholder "Provide a high-level description..."
Type what you want the code to do in plain English, then click ASK AI to generate the code automatically. You can also switch to CHAT to iterate on the output conversationally.

3. Pass input values to your code
Under Input Values, define the variables your code will use. Each variable gets a name (e.g., rawEmail) and a value mapped from a previous step (e.g., body.email).
Click ADD VARIABLE to add more inputs. These variables are directly accessible inside your code by their names.

4. View and edit the generated code
Click <> CODE to expand the code editor. This shows the JavaScript generated by Ask AI — you can review it, modify it, or write your own from scratch.


You don't have to use Ask AI. You can write your own JavaScript directly in the Code editor.
5. Test and save
Click TEST to run the step and verify the output. Once the result looks correct, click SAVE to commit the step to your workflow.
Real Use Cases
Split a full name into first and last name
Get current date and time in
YYYY-MM-DD HH:MM:SSformatCheck if a number is prime
Convert a string to uppercase
Extract the domain from a URL
Parse sender, subject, and body from a raw email string
Tips
Use
{{stepN.fieldname}}syntax to reference output from this step in later actionsLearn how to write better AI prompts: How to Write a Prompt
Supported Preinstalled Libraries
These libraries are available in the JS Code environment without any imports needed:
json
[...][ "form-data as FormData", "https", "crypto", "setTimeout", "axios", "jsonwebtoken as jwt", "lodash as _", "node-fetch as fetch", "cheerio", "moment", "fetch", "Buffer", "atob", "XMLParser", "XMLBuilder", "XMLValidator" ]