ChatBot is a universal and comprehensive framework for building modern AI chatbots. It can be used by total beginners as well as advanced developers.

ChatBot

This should return a list of dictionaries of content type JSON in reverse chronological order. For example: `[{"id":1234},{"id":1235}]`

If the returned JSON is not a list and is instead an object (maybe paginated), enter the key that contains the results. Example: "results", "items", "objects", etc... (children via dot syntax supported)

Zapier deduplicates the array we see each poll against the `id` key. If the `id` key does not exist, you should specify an alternative unique key to deduplicate off of. If neither are supplied, we fallback to looking for the shortest key with `id` in it otherwise we will raise an error.

If the content fetched is XML or HTML, optionally define an xpath to extract data from the content. Will trigger whenever the content at the xpath changes!

A pipe (`|`) separated username, password combo for standard HTTP authentication.

Key, value pairs to be added as headers in all requests.

By default, Zapier gives you the entire payload of the webhook. If this is specified, Zapier will only grab the child key from the object sent to Zapier. For example, given `{"contact": {"name": "Mike"}}`, specify `contact` to only trigger on `{"name": "Mike"}`. Traverse more deeply nested children by using dot-separated syntax.

By default, Zapier gives you the entire payload of the webhook. If this is specified, Zapier will only grab the child key from the object(s) sent to Zapier. For example, given `{"contact": {"name": "Mike"}}`, add "contact" here to only return `{"name": "Mike"}`. Traverse more deeply nested children by using dot-separated syntax. Line item payloads will run your Zap for each instance of the child key.

If the content fetched is HTML, optionally define an xpath to extract data from the content. Will trigger whenever the content at the xpath changes! (Does not support XML.)

Zapier deduplicates the array we see each poll against the `id` key. If the `id` key does not exist, you should specify an alternative unique key to deduplicate off of. If neither are supplied, we fallback to looking for the shortest key with `id` in it and finally a hash of every key and value.

The HTTP method we'll use to perform the request.

Any URL with a querystring will be re-encoded properly.

If yes, request `data` below will be overridden with the raw data from the previous step (you are still responsible for headers though, like `Content-Type`). Defaults to no.

Place raw data for the request here. Could be JSON, XML, etc. Nothing is parsed for you here (including extra line breaks or spaces -- everything is sent verbatim). You probably need to define a header below corresponding to your data `Content-Type`. Reminder: data provided in this field is completely ignored if you chose "yes" to pass-through.

By default we convert fields with double underscores into nested dictionaries. Select no here to turn this behavior off.

We recommend *not* putting any query string arguments in the URL (use "Query String Params" below).

These params will be URL-encoded and appended to the URL when making the request. *Note:* If you specify nothing for this field, **we will automatically encode and include every field** from the previous step in the query string. If you don't want this, use the "Custom Request" action.

If "yes", this dumps the results directly to a key (set below) as JSON. If "no", we simulate a form.

Pay special attention to the proper mapping of the data below.

These are key-value pairs that will form the data portion of the request. If you leave these empty, all fields from the previous step will be used. If you want to send your own raw JSON or form-encoded values here, please use the **Custom Request** action instead.

Wraps the request in an array if that is something the service requires.

A file object to be attached to the request. The request will automatically be converted to `multipart/form-data`, with the above payload type defining the sub content-type for the data being passed.

Webhooks simply POST data (or JSON) to a specific URL every time we see something new. Webhooks can also accept data (or JSON) to kick off a workflow in Zapier.

Webhooks by Zapier

Stay on top of incoming messages from your ChatBot app by setting up this workflow. As soon as you receive a new message in ChatBot, this automation will launch a GET request using Webhooks by Zapier. Not only does this streamline your online conversations, but it also ensures you're promptly responding to all customer queries and feedback.

ChatBot + Webhooks by Zapier

Fetch new ChatBot messages with gets in Webhooks by Zapier

Supported triggers and actions

New Message

Catch Raw Hook

Custom Request

POST

Retrieve Poll

Catch Hook

GET

PUT

About ChatBot

Related categories

About Webhooks by Zapier

Related categories