> ## Documentation Index
> Fetch the complete documentation index at: https://docs.coderabbit.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Triage support tickets

> Set up a CodeRabbit automation that investigates each new Zendesk, Intercom, or Pylon ticket against your code and docs and posts a summary, a classification, and a draft customer reply to Slack.

Cloud Only (This feature is available on CodeRabbit Cloud only. It is not available on self-hosted CodeRabbit deployments.)

## What you get

For each new support ticket, CodeRabbit summarizes the problem and customer impact, says whether it looks like a bug, expected behavior, a known issue, or needs more information, and cites docs, issues, and pull requests. It suggests severity and next steps and ends with a draft customer reply for a support engineer to review. It doesn't send the reply or change the ticket.

Pick the route by where tickets show up:

| Tickets show up | Start from |
| - | - |
| In a Slack channel, posted by the support tool's Slack app or escalated there by support engineers | [`/automations` in the channel](#tickets-already-post-to-a-slack-channel) |
| Only in the support tool | [A **Zendesk**, **Intercom**, or **Pylon** trigger](#trigger-on-new-tickets-in-the-support-tool) |

## Before you start

| Requirement | Who does it | Where |
| - | - | - |
| Your repositories and the support tool as connections, so CodeRabbit can read the whole ticket. Add Linear or Jira to find known issues. | A CodeRabbit admin | [Connections](https://docs.coderabbit.ai/connections.md) |
| The connections in the run's scope | A CodeRabbit admin | **Slack** > **Channel mappings** for the channel, or [Base Scope](https://docs.coderabbit.ai/automations/destinations.md#connections-and-base-scope) |
| CodeRabbit in the channel, if it's private | A channel member | `/invite @coderabbit` in the channel |
| For the support tool trigger: a webhook in the support tool | An admin of the support tool | Zendesk, Intercom, or Pylon |

Runs that post to Slack are billed like CodeRabbit Agent for Slack. See [Who can use Automations](https://docs.coderabbit.ai/automations.md#who-can-use-automations).

## Set it up

### Tickets already post to a Slack channel

In the channel, run `/automations` and start a new automation. In **Purpose**, choose **Investigate support tickets**, in **When to run**, choose **New channel message**, and select **Save and activate**. The automation is active right away, without a trial run, and runs on every new top-level message in the channel. To run only on ticket messages, see [Channel messages](https://docs.coderabbit.ai/automations/triggers.md#channel-messages).

### Trigger on new tickets in the support tool

**1. Create the automation**

Open **Automations** > **Home** and select **Create Automation**.

**2. Choose the trigger**

In the **Trigger** card, select **Zendesk**, **Intercom**, or **Pylon**, and choose the event: **Ticket created** for Zendesk, **Customer started conversation** for Intercom, or **Issue created** for Pylon.

**3. Write the agent instructions**

In **Agent instructions**, describe the triage, for example:

```text wrap theme={null}
Investigate the new support ticket using our repositories and connected tools. Summarize the problem and customer impact. Say whether it is a bug, expected behavior, a known issue, or needs more information, and cite docs, issues, and PRs. Suggest severity and next steps. End with a draft customer reply for review. Do not send it to the customer or change the ticket.
```

**4. Choose the destination**

In the **Target** card, choose the Slack channel where support engineers review the result. To also write to the ticket, select **Add target**, add the support tool's connection, and say what to write in its **Target instructions**.

**5. Save and connect the webhook**

Select **Save automation**. The editor shows the webhook URL and setup steps for your support tool; for Pylon, that's a Pylon trigger with a **Send webhook** action. Configure the webhook in the support tool. See [Other providers](https://docs.coderabbit.ai/automations/triggers.md#other-providers).

## Test it

Create a test ticket in the support tool, or post a test ticket message in the channel. See [Test a trigger](https://docs.coderabbit.ai/automations/triggers.md#test-a-trigger).

## Check the runs

Each run is listed on the automation's page. See [Automation details](https://docs.coderabbit.ai/automations/manage.md#automation-details). A successful run posts the summary, classification, severity, next steps, and draft reply as a reply in the ticket message's thread, or as a new thread in the destination channel for the support tool trigger. If you added a support tool target, check the ticket for what its **Target instructions** asked for.

If no run appears for a support tool event, check that the automation is active, that the event type matches, and that the webhook in the support tool points at the URL from the editor.

## What's next

> [card] [Triage new engineering tickets](https://docs.coderabbit.ai/automations/guides/triage-engineering-tickets.md)
>
> Investigate new Linear, Jira, or GitHub issues against your code.

> [card] [Automation destinations](https://docs.coderabbit.ai/automations/destinations.md)
>
> Write results back to the ticket as well as to Slack.

> [card] [Knowledge Base in Slack](https://docs.coderabbit.ai/slack-agent/knowledge-base.md)
>
> Teach CodeRabbit product facts so triage gets them right.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.