# How to use webhooks in auto comment tool?

**URL:** <https://forum.ideta.io/t/how-to-use-webhooks-in-auto-comment-tool/178>\
**Category:** Community Management Tool/  Auto Comment & Reply\
**Created:** [December 28, 2022, 10:16am UTC](https://forum.ideta.io/t/how-to-use-webhooks-in-auto-comment-tool/178 "2022-12-28T10:16:12Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Glen](https://avatars.discourse-cdn.com/v4/letter/g/8c91f0/32.png) [@Glen](https://forum.ideta.io/u/Glen)\
**Post date:** [December 28, 2022, 10:16am UTC](https://forum.ideta.io/t/how-to-use-webhooks-in-auto-comment-tool/178/1 "2022-12-28T10:16:12Z")

</div>

How can I build a webhook for auto comment ? What is the expected response format ?

---

<div class="post-metadata">

**Author:** ![Glen](https://avatars.discourse-cdn.com/v4/letter/g/8c91f0/32.png) [@Glen](https://forum.ideta.io/u/Glen)\
**Post date:** [January 11, 2023, 1:10pm UTC](https://forum.ideta.io/t/how-to-use-webhooks-in-auto-comment-tool/178/2 "2023-01-11T13:10:05Z")

</div>

You can either build your own webhook using code or you can use our webhook creation tool.

If you use our webhook creation tool, this is going to be easier since we have templates you can use.

Go to [https://webhooks.ideta.io](https://connect.ideta.io?rd=https://webhooks.ideta.io)

Then click on “Create a webhook” and follow the steps :

 ![Capture d’écran 2023-01-11 à 13.36.14](https://europe1.discourse-cdn.com/flex017/uploads/ideta/original/1X/89e04e8301f4faa4f3a489b07346668f6502e288.jpeg)

Make sure to select the Auto Comment Template :

 ![Capture d’écran 2023-01-11 à 13.42.25](https://europe1.discourse-cdn.com/flex017/uploads/ideta/original/1X/ae4b6595c748cf7259b44042223957cff4725dda.png)

The “Extract comment value” bubble has little interest.

The “Filter” bubble will detect if the comment contains a given value :

 ![correctionimage](https://europe1.discourse-cdn.com/flex017/uploads/ideta/original/1X/ace553ec619566966027ac740a780d1bb405cf94.png)

You can set up multiple answers or have multiple filter step one after the other. You just have to create them.

Then you can copy the webhook URL in the Publish section (make sure it is published) :

 ![Capture d’écran 2023-01-11 à 13.52.40](https://europe1.discourse-cdn.com/flex017/uploads/ideta/original/1X/0a0ec43057129ac148c3ebeac4c400500d189c17.png)

In both cases (code or webhooks tool), the output has to be JSON and in the following format (see Answer bubble) :

 ![Capture d’écran 2023-01-11 à 13.34.22](https://europe1.discourse-cdn.com/flex017/uploads/ideta/original/1X/677b37633179a46672a75ff6499658fed8bcf912.png)

The generic format is the following :  
{  
“reactions”: string[] | string,  
“privateMessages”:string[] | string,  
“publicMessages”: string[] | string  
}

For example, this can look like that :

{  
“reactions”: [“like”, “angry”],  
“privateMessages”:[“answer 1”, “answer 2”],  
“publicMessages”:[“answer 3”, “answer 4”]  
}

or like that :

{  
“reactions”: “like”,  
“privateMessages”:[“answer 1”, “answer 2”],  
“publicMessages”: “answer3”  
}

Then you can use it in the auto comment tool :

 ![Capture d’écran 2023-01-11 à 14.01.36](https://europe1.discourse-cdn.com/flex017/uploads/ideta/original/1X/480935b8923a3e5c5f1aea3088110e4d57728e03.png)

This is just the basics ! We have Natural Language Processing and Sentiment Analysis available. And you can combine everything in a complex workflow.

Do not hesitate to contact [support@ideta.io](mailto:support@ideta.io) to get additional help.