# POST /api/notifications/deliveries/outcome-summary

Counts how deliveries ended in a time window, broken down by outcome.

This is drop accounting for an outage: how many events were still queued, how many were
skipped without an outbound attempt because the destination was suppressed, and how many
exhausted their retries and were dead-lettered, over the period a receiver was unreachable.
It counts in the store and reads no delivery, so the answer costs the same whether the window
holds ten deliveries or ten thousand.



The window is bounded, and a wider one is refused rather than quietly narrowed: an answer
silently covering a quarter of what was asked for would look exactly like a merchant who sent
nothing for the rest of it. Count a longer period in steps.



A POST because the window and its narrowing are a request body, matching the other
continuation and summary reads in this platform. Nothing is created or changed.

**Operation ID:** `notificationsGetDeliveryOutcomeSummary`

## Authorization

Requires: Notifications.Deliveries, merchant scope.

Requires API key scope: `webhooks`

Required permissions:
- `Notifications.Deliveries`

## Parameters

| Name | In | Required | Type | Description |
| --- | --- | --- | --- | --- |
| suppressNulls | query | no | boolean | If true, omit properties with null values. |

## Request Body

The window, and optionally the merchant and event type to narrow to.

**Content type:** `application/json`

Schema: [`GetNotificationDeliveryOutcomeSummaryInput`](https://devportal.demo.winkpg.io/docs/api/models/GetNotificationDeliveryOutcomeSummaryInput.md)

Properties:
- `fromUtc` (string(date-time)) required: Start of the window, inclusive, in UTC.
- `toUtc` (string(date-time)) required: End of the window, exclusive, in UTC.
- `merchantId` (string(uuid)): Count only this merchant's deliveries, or every merchant the caller's operating context  already reaches when omitted. (nullable)
- `eventType` (string): Count only this exact event type, or every type when omitted. (nullable, max length 512)

**Content type:** `text/json`

Schema: [`GetNotificationDeliveryOutcomeSummaryInput`](https://devportal.demo.winkpg.io/docs/api/models/GetNotificationDeliveryOutcomeSummaryInput.md)

Properties:
- `fromUtc` (string(date-time)) required: Start of the window, inclusive, in UTC.
- `toUtc` (string(date-time)) required: End of the window, exclusive, in UTC.
- `merchantId` (string(uuid)): Count only this merchant's deliveries, or every merchant the caller's operating context  already reaches when omitted. (nullable)
- `eventType` (string): Count only this exact event type, or every type when omitted. (nullable, max length 512)

**Content type:** `application/*+json`

Schema: [`GetNotificationDeliveryOutcomeSummaryInput`](https://devportal.demo.winkpg.io/docs/api/models/GetNotificationDeliveryOutcomeSummaryInput.md)

Properties:
- `fromUtc` (string(date-time)) required: Start of the window, inclusive, in UTC.
- `toUtc` (string(date-time)) required: End of the window, exclusive, in UTC.
- `merchantId` (string(uuid)): Count only this merchant's deliveries, or every merchant the caller's operating context  already reaches when omitted. (nullable)
- `eventType` (string): Count only this exact event type, or every type when omitted. (nullable, max length 512)

## Responses

### 200

The count of deliveries in each state, plus the window they cover.

**Content type:** `application/json`

Schema: [`NotificationDeliveryOutcomeSummaryDto`](https://devportal.demo.winkpg.io/docs/api/models/NotificationDeliveryOutcomeSummaryDto.md)

Properties:
- `fromUtc` (string(date-time)): Start of the counted window, inclusive, in UTC.
- `toUtc` (string(date-time)): End of the counted window, exclusive, in UTC.
- `pending` (integer(int64)): Deliveries created but not yet picked up by a processor: still queued on this side.
- `processing` (integer(int64)): Deliveries in flight when the window was counted.
- `succeeded` (integer(int64)): Deliveries the destination acknowledged. The "delivered" number.
- `failed` (integer(int64)): Deliveries that failed an attempt and may still be retried.
- `deadLettered` (integer(int64)): Deliveries that exhausted every retry and were dead-lettered.
- `skipped` (integer(int64)): Deliveries never attempted, because the destination was suppressed or another policy gate  stopped them. The closest thing this platform has to "discarded".
- `total` (integer(int64)): Every delivery in the window, in any state. (read only)

### 400

The window is missing, inverted, or wider than the bound.

**Content type:** `application/json`

Schema: [`RemoteServiceErrorResponse`](https://devportal.demo.winkpg.io/docs/api/models/RemoteServiceErrorResponse.md)

Properties:
- `error` ([RemoteServiceErrorInfo](https://devportal.demo.winkpg.io/docs/api/models/RemoteServiceErrorInfo.md))

_Example: Request failed validation_

One or more fields were rejected before any rule ran. error.code is null on the envelope and the causes are in validationErrors, one entry per failure. An entry carries code when the validator declared a publishable one, and memberPaths when the full path says more than members does (an element inside a collection, for example). Both keys are omitted when they would add nothing.

```json
{
  "error": {
    "code": "null",
    "message": "Your request is not valid!",
    "details": "The following errors were detected during validation.",
    "data": {
      "correlationId": "0123456789abcdef0123456789abcdef"
    },
    "validationErrors": [
      {
        "message": "'Amount' must be greater than 0.",
        "members": [
          "amount"
        ]
      },
      {
        "message": "The card has expired.",
        "members": [
          "expirationDate"
        ],
        "code": "CardData:CardExpired"
      },
      {
        "message": "'Value' must not be empty.",
        "members": [
          "value"
        ],
        "memberPaths": [
          "customFields[2].value"
        ]
      }
    ]
  }
}
```

### 403

The caller may not read delivery history.

### default

The request failed. The body carries the standard error envelope: a machine-readable `error.code`, a human-readable `error.message`, and `error.validationErrors` when the failure was a validation rejection. The [error reference](https://devportal.demo.winkpg.io/docs/errors.md) lists the values `error.code` can take.

**Content type:** `application/json`

Schema: [`RemoteServiceErrorResponse`](https://devportal.demo.winkpg.io/docs/api/models/RemoteServiceErrorResponse.md)

Properties:
- `error` ([RemoteServiceErrorInfo](https://devportal.demo.winkpg.io/docs/api/models/RemoteServiceErrorInfo.md))

_Example: Business rule refused the request_

The request was understood and rejected by a platform rule. This is the shape that carries a code from the catalog above, so branch on error.code rather than on the message text. The HTTP status is the one listed against the code.

```json
{
  "error": {
    "code": "Transactions:OperationNotAllowed",
    "message": "The operation 'Void' is not allowed on a Sale transaction with result 'Approved' and settlement state 'Settled'.",
    "details": "null",
    "data": {
      "correlationId": "0123456789abcdef0123456789abcdef"
    },
    "validationErrors": "null"
  }
}
```

_Example: Resource not found_

The id in the path names no record the caller can see. Most resources answer with a null error.code and a message naming the resource type and the id; a resource that publishes a not-found code in the catalog above carries it in error.code instead, so check for one before matching on the message.

```json
{
  "error": {
    "code": "null",
    "message": "There is no entity Transaction with id = 7d1f3a52-0c4e-4b8a-9e21-5a6b7c8d9e00!",
    "details": "null",
    "data": {
      "correlationId": "0123456789abcdef0123456789abcdef"
    },
    "validationErrors": "null"
  }
}
```

_Example: Unexpected server fault_

Something failed that no rule anticipated. There is no code to branch on and the message is deliberately generic: the diagnostic detail stays in the platform's logs rather than going to the caller. Quote error.data.correlationId (the same value as the X-Correlation-Id response header) when you contact support: it's what finds the failure in those logs. Retrying an unsafe operation after this risks a duplicate, so reconcile before you retry.

```json
{
  "error": {
    "code": "null",
    "message": "An internal error occurred during your request!",
    "details": "null",
    "data": {
      "correlationId": "0123456789abcdef0123456789abcdef"
    },
    "validationErrors": "null"
  }
}
```

### 429

The request was refused because a rate limit was exceeded, or because something a later retry can clear stopped it. A rate limit refusal carries an `application/problem+json` body: wait at least the interval `Retry-After` names before retrying, then back off. Limits are tuned per deployment, so read the allowance from the response headers rather than assuming a fixed ceiling. Any other refusal carries the standard error envelope as `application/json`, and its `error.code` names the cause.

**Content type:** `application/problem+json`

Schema: [`RateLimitProblemDetails`](https://devportal.demo.winkpg.io/docs/api/models/RateLimitProblemDetails.md)

Properties:
- `type` (string) required: The problem type identifier. Always the same value: the failure is the status code itself,  so there is no sub-type for a caller to branch on. (nullable)
- `title` (string) required: A short, human-readable summary of the problem type. (nullable)
- `status` (integer(int32)) required: The HTTP status code, repeated in the body as the problem-details format defines.
- `detail` (string) required: A human-readable explanation of this occurrence of the problem. (nullable)
- `retryAfterSeconds` (integer(int32)) required: How long to wait before retrying, in whole seconds, carrying the same figure as the  `Retry-After` header. Always at least one: a value of zero would invite an immediate  retry that is certain to be rejected again.

**Content type:** `application/json`

Schema: [`RemoteServiceErrorResponse`](https://devportal.demo.winkpg.io/docs/api/models/RemoteServiceErrorResponse.md)

Properties:
- `error` ([RemoteServiceErrorInfo](https://devportal.demo.winkpg.io/docs/api/models/RemoteServiceErrorInfo.md))

## Example request

Every block below sends the same request. Replace {{BASE_URL}} with the address of the API you are calling and {{API_KEY}} with your own key.

The request body is a GetNotificationDeliveryOutcomeSummaryInput. See the Request body section below for its fields.

### cURL

```bash
curl -X POST "{{BASE_URL}}/api/notifications/deliveries/outcome-summary" \
  -H "api-key: {{API_KEY}}" \
  -H "Content-Type: application/json" \
  -d '{
  "fromUtc": "2026-01-01T00:00:00Z",
  "toUtc": "2026-01-01T00:00:00Z"
}'
```

### PowerShell

```powershell
$headers = @{
    'api-key' = '{{API_KEY}}'
}

$body = @'
{
  "fromUtc": "2026-01-01T00:00:00Z",
  "toUtc": "2026-01-01T00:00:00Z"
}
'@

$response = Invoke-RestMethod -Method POST -Uri '{{BASE_URL}}/api/notifications/deliveries/outcome-summary' `
    -Headers $headers -ContentType 'application/json' -Body $body
```

### TypeScript (SDK)

```bash
npm install @winkpg/winkpg-api
```

```typescript
import { Configuration, NotificationsApi } from '@winkpg/winkpg-api';

const api = new NotificationsApi(new Configuration({
  basePath: '{{BASE_URL}}',
  apiKey: '{{API_KEY}}',
}));

const { data } = await api.notificationsGetDeliveryOutcomeSummary({
  "fromUtc": "2026-01-01T00:00:00Z",
  "toUtc": "2026-01-01T00:00:00Z"
});
```

### TypeScript (raw HTTP)

```typescript
const response = await fetch('{{BASE_URL}}/api/notifications/deliveries/outcome-summary', {
  method: 'POST',
  headers: {
    "api-key": "{{API_KEY}}",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "fromUtc": "2026-01-01T00:00:00Z",
    "toUtc": "2026-01-01T00:00:00Z"
  }),
});

const data = await response.json();
```

### C# (SDK)

```bash
dotnet add package WinkPg.Api.Client
```

```csharp
using WinkPg.Api.Client.Api;
using WinkPg.Api.Client.Client;
using System.Text.Json;

var config = new Configuration { BasePath = "{{BASE_URL}}" };
config.AddApiKey("api-key", "{{API_KEY}}");

var api = new NotificationsApi(config);
var body = JsonSerializer.Deserialize<GetNotificationDeliveryOutcomeSummaryInput>("""
    {
      "fromUtc": "2026-01-01T00:00:00Z",
      "toUtc": "2026-01-01T00:00:00Z"
    }
    """);

var result = await api.NotificationsGetDeliveryOutcomeSummaryAsync(body);
```

### C# (raw HTTP)

```csharp
using System.Text;

using var http = new HttpClient { BaseAddress = new Uri("{{BASE_URL}}") };

var request = new HttpRequestMessage(new HttpMethod("POST"), "/api/notifications/deliveries/outcome-summary");
request.Headers.Add("api-key", "{{API_KEY}}");

request.Content = new StringContent("""
    {
      "fromUtc": "2026-01-01T00:00:00Z",
      "toUtc": "2026-01-01T00:00:00Z"
    }
    """, Encoding.UTF8, "application/json");

var response = await http.SendAsync(request);
response.EnsureSuccessStatusCode();
var json = await response.Content.ReadAsStringAsync();
```

### Python (SDK)

```bash
pip install winkpg-api
```

```python
import winkpg_api

configuration = winkpg_api.Configuration(host="{{BASE_URL}}")
configuration.api_key["ApiKey"] = "{{API_KEY}}"

with winkpg_api.ApiClient(configuration) as client:
    api = winkpg_api.NotificationsApi(client)
    body = winkpg_api.GetNotificationDeliveryOutcomeSummaryInput.from_dict({
      "fromUtc": "2026-01-01T00:00:00Z",
      "toUtc": "2026-01-01T00:00:00Z"
    })
    result = api.notifications_get_delivery_outcome_summary(body)
```

### Python (raw HTTP)

```bash
pip install requests
```

```python
import requests

headers = {
    "api-key": "{{API_KEY}}",
    "Content-Type": "application/json",
}

body = {
  "fromUtc": "2026-01-01T00:00:00Z",
  "toUtc": "2026-01-01T00:00:00Z"
}

response = requests.request(
    "POST",
    "{{BASE_URL}}/api/notifications/deliveries/outcome-summary",
    headers=headers,
    json=body,
)
response.raise_for_status()
data = response.json()
```

## See also

- [All documentation](https://devportal.demo.winkpg.io/llms.txt): the machine-readable index of every public page on this site.
