> For the complete documentation index, see [llms.txt](https://developer.harness.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.harness.io/ai-sre/ai-sre-for-incident-responders/create-incidents.md).

# Create Incidents

When you spot a problem, whether from a customer report, degraded performance, or a teammate flagging something, you can declare an incident in Harness AI SRE.

The AI-powered Quick Start feature analyzes your problem description and auto-populates incident fields, so you can get an incident up and running in seconds.

### Create incidents from the UI <a href="#create-incidents-from-the-ui" id="create-incidents-from-the-ui"></a>

{% tabs %}
{% tab title="Step by Step" %}

#### Step 1: Access incident creation <a href="#step-1-access-incident-creation" id="step-1-access-incident-creation"></a>

Open the incident list and start a new incident:

1. Navigate to **Incidents** from the left panel.

   ![Incident list](/files/zAG2l3DdRwx6OE88974W)
2. Choose your creation method:
   * Click **New Incident** for a blank incident.
   * Select an **Incident Type** from the dropdown next to "New Incident" to use a pre-configured template.

#### Step 2: Select incident type <a href="#step-2-select-incident-type" id="step-2-select-incident-type"></a>

Pick the incident type that matches the problem:

1. Choose the appropriate **Incident Type** from the available options.

   <div data-with-frame="true"><figure><img src="/files/lhWPS6oSIhVlvkQMj13b" alt="Create new incident form"><figcaption></figcaption></figure></div>
2. This will pre-populate relevant fields and associate appropriate runbooks.
3. If you are unsure which type to pick, check with your team lead, incident types are configured by your administrator to match your organization's response processes.

#### Step 3: Describe the problem <a href="#step-3-describe-the-problem" id="step-3-describe-the-problem"></a>

Describe the problem so the AI can suggest field values:

1. Use the **Quick Start** block to describe the problem in plain language.
2. Provide a clear, concise description including affected services, symptoms, and impact.
3. The AI system will analyze your description and suggest field values automatically.

#### Step 4: Generate incident fields <a href="#step-4-generate-incident-fields" id="step-4-generate-incident-fields"></a>

Run the AI analysis to populate the incident fields:

1. Click the **up arrow** to process your description.
2. AI will automatically populate incident fields (severity, priority, etc.) based on what you wrote.

#### Step 5: Review and customize <a href="#step-5-review-and-customize" id="step-5-review-and-customize"></a>

Confirm the suggested values and save the incident:

1. Review all auto-generated field values.
2. Manually adjust anything that needs correction:
   * Title and description
   * Severity and priority levels
   * Assignee
   * Any custom fields specific to the incident type
3. Click **Save** to create the incident.
   {% endtab %}

{% tab title="Interactive Guide" %}
{% embed url="<https://app.tango.us/app/embed/500fd8d9-da9a-42b3-8006-94ecc9d5d8ff?skipCover=true&defaultListView=false&skipBranding=false&makeViewOnly=false&hideAuthorAndDetails=true>" %}
Create an Incident
{% endembed %}

Follow this interactive guide to create an incident with AI-powered assistance.
{% endtab %}
{% endtabs %}

***

### Create incidents from Slack <a href="#create-incidents-from-slack" id="create-incidents-from-slack"></a>

You can also create incidents directly from Slack using the `/harness new` command. This is especially useful during on-call shifts when you want to stay in your communication tool.

![Slack incident notification](/files/9duoKq27h6vBeWgF7MMU)

Go to [Managing incidents in Slack](/ai-sre/ai-sre-for-incident-responders/slack-commands.md) to review the full command reference.

***

### Best practices for incident creation <a href="#best-practices-for-incident-creation" id="best-practices-for-incident-creation"></a>

Follow these practices to create accurate, actionable incidents:

* **Choose the right incident type:** Select the most specific type available to get the right runbooks and field configuration.
* **Write a clear description:** Good Quick Start descriptions lead to more accurate AI field suggestions. Include what is affected, what you are observing, and when it started.
* **Verify auto-generated fields:** Always review and adjust AI-suggested values to ensure accuracy.
* **Set correct severity levels:** Align severity with actual business impact and response time requirements.

***

### What happens next <a href="#what-happens-next" id="what-happens-next"></a>

Once you save the incident:

* An **incident channel** is automatically created in your communication tool (Slack or Teams).
* **Relevant team members** are notified based on the incident type's configuration.
* Any **runbooks** associated with the incident type may be auto-attached.
* The **Incident Details** page opens, where you manage the incident through resolution.

Go to [Active incident management](/ai-sre/ai-sre-for-incident-responders/respond-to-active-incidents/manage-incidents.md) to work through an active incident.

***

### Next steps <a href="#next-steps" id="next-steps"></a>

* Go to [Active incident management](/ai-sre/ai-sre-for-incident-responders/respond-to-active-incidents/manage-incidents.md) to work through an active incident.
* Go to [Create and track action items](/ai-sre/ai-sre-for-incident-responders/respond-to-active-incidents/create-and-track-action-items.md) to assign follow-up work.
* Go to [Managing incidents in Slack](/ai-sre/ai-sre-for-incident-responders/slack-commands.md) to run incidents from Slack.