> 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/api-reference/more-products-apis/ai-sre.md).

# AI SRE

Manage incidents, alerts, deploys, and activity timelines with the Harness AI SRE API.

The AI SRE API manages the incidents, alerts, and deploys that AI SRE tracks, including root-cause theories and activity timelines. This reference covers the REST endpoints available today. Not every AI SRE capability is on the API yet.

## Request basics

* **API key token**: Send it in the `x-api-key` header on every request. `Authorization: Bearer` is the browser-session path and rejects API keys. Use a service account token for automation. Go to [Manage API keys](https://developer.harness.io/harness-platform/use-harness-platform/automation/api/add-and-manage-api-keys) to create a key and token.
* **Account, organization, and project IDs**: Send `accountId`, `orgId`, and `projectId` as query parameters. All three are required on every endpoint. Go to [Organizations and projects](https://developer.harness.io/harness-platform/use-harness-platform/organizations-and-projects) to find these identifiers.
* **Base URL**: Send requests to `https://app.harness.io/gateway/ir/tp`, followed by the endpoint path.

## Example request

List the open incidents in a project:

```bash
curl "https://app.harness.io/gateway/ir/tp/api/v1/mc/incidents?accountId=ACCOUNT_ID&orgId=ORG_ID&projectId=PROJECT_ID&status=investigating&pageSize=10&page=0" \
  -H "x-api-key: $HARNESS_API_KEY"
```

Replace `ACCOUNT_ID`, `ORG_ID`, and `PROJECT_ID` with your own values, and set `HARNESS_API_KEY` to your API key token.

## Find an endpoint

Endpoint paths are relative to the base URL.

* **Incidents**: Create an incident with `POST /api/v1/mc/incidents`, list them with `GET /api/v1/mc/incidents`, and read or update one with `GET` and `PATCH /api/v1/mc/incidents/{incidentId}`. Close an incident with `POST /api/v1/mc/incidents/{incidentId}/close`.
* **Root-cause theories**: Add a theory to an incident with `POST /api/v1/mc/incidents/{incidentId}/root-cause-theories`. Then confirm it, rule it out, or undo that decision with `POST` on `/confirm`, `/rule-out`, or `/undo` under `/api/v1/mc/incidents/{incidentId}/root-cause-theories/{theoryId}`.
* **Alerts**: List alerts with `GET /api/v1/mc/alerts` and read or update one with `GET` and `PATCH /api/v1/mc/alerts/{alertId}`. Change an alert's state with `POST` on `/acknowledge`, `/dismiss`, or `/resolve` under `/api/v1/mc/alerts/{alertId}`.
* **Deploys**: List deploys with `GET /api/v1/mc/deploys` and read one with `GET /api/v1/mc/deploys/{deployId}`.
* **Activity timelines**: Read an activity's timeline with `GET /api/v1/mc/activities/{activityId}/timeline` and add a markdown note with `POST` to the same path.

## Known behavior

* **Time filters**: `reportedAfter` and `reportedBefore` take ISO-8601 values such as `2026-05-01T00:00:00Z`, not epoch milliseconds.
* **Pagination**: `page` is 0-based. Use a `pageSize` of 20 or less, because larger page sizes can return a 500 error.
