Receive data in Mammoth with a webhook
A webhook gives a dataset its own URL that other systems send data to whenever something happens, instead of Mammoth pulling data on a schedule. In the Data Library, click New dataset and choose Webhook, or open the Connectors launcher and choose Push, then Webhook. The dataset's Webhook card holds the URL, the secret and a log of what arrived.
Use a webhook when an external platform supports outbound webhooks (a payment platform, a form tool, a CRM or your own app), when your source sends data on events (a new order, a form submission, a status change) rather than making it available to read, or when another system sends data as one step of a larger workflow. Use a connector when Mammoth should pull from a database or API, and Upload files when data arrives as files.
Create a webhook dataset
- Go to the Data Library and click New dataset.
- Choose Webhook ("Generate a URL to receive data from other applications").
- Enter the New dataset name (the box reads "New webhook name").
- Under Add new rows, pick how often arriving rows are added: 5 min (selected first), Hourly, Daily or By hand. You can change this later on the card.
- Click Create new dataset.
Mammoth opens the dataset on its Webhook card. As the window says, "Other systems POST JSON to it; each object’s keys become columns. Its URL and secret are on the dataset once it is made." A webhook made this way adds new rows to the rows already there.
From the Connectors launcher, choose Push, then Webhook, or type "webhook" in the search box. That opens the connector workspace, whose chat starts with "Create a webhook that adds the JSON it receives to a new dataset." and asks for details such as the dataset name.

Find the webhook URL and secret
Open the dataset's panel in the Data Library. The Webhook card shows everything a sender needs:
| Item | What it is |
|---|---|
| URL | The address your system sends data to. Click Copy URL. |
| Secret | The value your system must send in an X-SECRET request header. Click Copy secret. With no secret, the card says "None. Anyone with the URL can add rows." (see Secure a webhook) |
| Copy curl | Copies a ready-to-run curl command that posts a sample JSON row to the URL, with the X-SECRET header when there is a secret |
Paste the URL, and the secret if there is one, into your external system's webhook settings. The URL looks like https://app.mammoth.io/api/v1/webhook/data/7fK2pQ9xLm4Rt8Vw3Yb6Nc1D, where the key at the end is 24 random letters and digits. The key is the only thing that identifies the webhook, so treat the URL like a password when the webhook has no secret.
Send data to the webhook URL
Mammoth accepts:
- POST with a JSON body (
Content-Type: application/json); - POST with form fields, which are stored as JSON;
- GET with query parameters, which are stored as JSON.
Each request becomes one row, so send one JSON object per request: a JSON array arrives as a single row of raw JSON. Only a Content-Type of application/json (a charset parameter is fine) is read as JSON; any other body is read as form fields.
Until the first row arrives, the card shows "Waiting for the first row. Send one JSON object per row; its keys become the columns." with a sample curl command. To check the setup without your source system, click Send a test row while the card is waiting. Mammoth shows "Test row sent. It is being added now."
Choose when new rows are added
Rows that arrive wait until they are added to the dataset. Pick how often under Add new rows:
| Option | When rows are added |
|---|---|
| 5 min | Every 5 minutes |
| Hourly | Every hour |
| Daily | Once a day. Used when no choice was saved. |
| By hand | Only when you click Add now |
While rows are waiting, the card says when, for example "New rows are waiting. They are added within the hour." Click Add now to add them straight away. If the card says "Each add replaces the rows already here with the new ones.", each add replaces the dataset's rows instead of adding to them.
Turn JSON keys into columns
Turn JSON keys into columns gives each top-level JSON key its own column. It is on for a new webhook, and a new key in a later row adds a column. When you turn it on, columns are added from the keys in the latest 500 rows already sent ("Saved. Columns are added from the rows already sent.").
With it off, each payload is stored as raw JSON that you can parse in a View. See Work with text and JSON.
Send data out of Mammoth instead
Mammoth has no outbound event webhooks: nothing calls your system when a pipeline finishes. To push a View's rows to your own HTTP endpoint after each run, use REST API Export (see Send data to a database or BI tool). To find out when a run is done, poll the API as described in API workflows.
Limits
| Limit | Value |
|---|---|
| Rows per request | One. A JSON array arrives as one row of raw JSON |
| Keys that become columns | Up to 200; a key longer than 100 characters is skipped |
| Rows read when you turn on Turn JSON keys into columns | The latest 500 rows already sent |
| Recent deliveries | The latest 10 posts |
| Secret length | 12 characters (see Secure a webhook) |
Troubleshooting webhooks
| Symptom | Cause | Fix |
|---|---|---|
| No data arriving | Your system sends to the wrong URL | Check it sends to the exact URL on the card |
| Recent deliveries shows Turned away | Wrong or missing X-SECRET, or a site the webhook doesn't allow | Fix the reason shown: send the secret from the card in an X-SECRET header (see Secure a webhook) |
| Posts show Received but the dataset hasn't changed | The rows are waiting to be added | Click Add now, or wait for the Add new rows schedule |
| Nothing shows in Recent deliveries | Your system isn't sending | Check your system's own delivery log or test button |
| Data looks wrong | Your system changed its payload | Check the latest rows. New keys add columns when Turn JSON keys into columns is on, so your pipeline may need updating for the new shape. |
Related
- Secure a webhook and check deliveries: the secret, allowed sites and Recent deliveries.
- Get data into Mammoth: compare webhooks with other ways in.
- Schedules and triggers: run automations when new data arrives.
- Batches: see and fix what arrived: how added rows are stored.
Something wrong? support@mammoth.io