zulip/templates/zerver/api/outgoing-webhooks.md

Outgoing Webhooks

Outgoing Webhooks allow you to build or set up Zulip integrations which are notified when certain types of messages are sent in Zulip. When one of those events is triggered, we'll send a HTTP POST payload to the webhook's configured URL. Webhooks can be used to power a wide range of Zulip integrations. For example, the Zulip Botserver is built on top of this API.

Zulip supports outgoing webhooks both in a clean native Zulip format, as well as a format that's compatible with Slack's outgoing webhook API, which can help with porting an existing Slack integration to work with Zulip.

To register an outgoing webhook:

• Log in to the Zulip server.
• Navigate to Settings () -> Your bots -> Add a new bot. Select Outgoing webhook for bot type, the URL you'd like Zulip to post to as the Endpoint URL, the format you want, and click on Create bot. to submit the form/
• Your new bot user will appear in the Active bots panel, which you can use to edit the bot's settings.

Triggering

There are currently two ways to trigger an outgoing webhook:

1. @-mention the bot user in a stream. If the bot replies, its reply will be sent to that stream and topic.
2. Send a private message with the bot as one of the recipients. If the bot replies, its reply will be sent to that thread.

Zulip message format

The Zulip-format webhook messages post the following data, encoded as JSON:

Name	Description
bot_email	Email of the bot user
data	The content of the message (in Markdown)
message	A dict containing details on the message which triggered the outgoing webhook
token	A string of alphanumeric characters you can use to authenticate the webhook request (each bot user uses a fixed token)
trigger	Trigger method

Some of the important fields in the message dict include the following:

Name	Description
recipient_id	Unique ID of the stream that will persist even if the stream is renamed
rendered_content	The content of the message, rendered in HTML

A correctly implemented endpoint will do the following:

• It will calculate a response that we call the "content" of the response.
• It will encode the content in Zulip's flavor of markdown (or just plain text).
• It will then make a dictionary with key of "content" and the value being that content. (Note that "response_string" is a deprecated synonym for "content".)
• It will encode that dictionary as JSON.

If the bot code wants to opt out of responding, it can explicitly encode a JSON dictionary that contains response_not_required set to True, so that no response message is sent to the user. (This is helpful to distinguish deliberate non-responses from bugs.)

Example incoming payload

{generate_code_example|zulip-outgoing-webhook-payload|fixture}

Slack-format webhook format

This interface translates the Zulip's outgoing webhook's request into Slack's outgoing webhook request. Hence the outgoing webhook bot would be able to post data to URLs which support Slack's outgoing webhooks. Here's how we fill in the fields that a Slack format webhook expects:

Name	Description
token	A string of alphanumeric characters you can use to authenticate the webhook request (each bot user uses a fixed token)
team_id	String ID of the Zulip organization
team_domain	Domain of the Zulip organization
channel_id	Stream ID
channel_name	Stream name
timestamp	Timestamp for when message was sent
user_id	ID of the user who sent the message
user_name	Full name of sender
text	The content of the message (in Markdown)
trigger_word	Trigger method
service_id	ID of the bot user

The above data is posted as list of tuples (not JSON), here's an example:

[('token', 'v9fpCdldZIej2bco3uoUvGp06PowKFOf'),
 ('team_id', 'zulip'),
 ('team_domain', 'zulip.com'),
 ('channel_id', '123'),
 ('channel_name', 'integrations'),
 ('timestamp', 1532078950),
 ('user_id', 21),
 ('user_name', 'Sample User'),
 ('text', '@**test**'),
 ('trigger_word', 'mention'),
 ('service_id', 27)]

• For successful request, if data is returned, it returns that data, else it returns a blank response.
• For failed request, it returns the reason of failure, as returned by the server, or the exception message.