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

# Response dictionary

> Field-by-field definitions for prohibited, safe, invalid, and summary results.

A successful scrub returns HTTP `200` with four top-level fields:

```json theme={null}
{
  "prohibited_numbers_to_call": ["+52 55 1234 5678"],
  "safe_numbers_to_call": ["55 8765 4321"],
  "invalid_numbers": [
    {
      "input": "abc",
      "reason": "unparseable"
    }
  ],
  "summary": {
    "total_checked": 3,
    "prohibited_numbers": 1,
    "safe_numbers": 1,
    "invalid_numbers": 1
  }
}
```

## Top-level fields

| Field                        | Type             | Meaning                                                                                                               |
| ---------------------------- | ---------------- | --------------------------------------------------------------------------------------------------------------------- |
| `prohibited_numbers_to_call` | `string[]`       | Original, structurally valid inputs found in the selected DNC registry. Do not call these numbers.                    |
| `safe_numbers_to_call`       | `string[]`       | Original, structurally valid inputs not found in the selected registry at the time of the check.                      |
| `invalid_numbers`            | `InvalidPhone[]` | Inputs that could not be normalized for the selected country. Each item contains the original `input` and a `reason`. |
| `summary`                    | `object`         | A complete count of prohibited, safe, and invalid results.                                                            |

<Warning>
  In `safe_numbers_to_call`, “safe” means only “not found in the selected DNC
  registry at the time of this check.” It does not prove that the number is
  active, reachable, assigned, supported by consent, or callable under every
  other law or rule.
</Warning>

## Summary

| Field                        | Meaning                                                                                |
| ---------------------------- | -------------------------------------------------------------------------------------- |
| `summary.total_checked`      | Total array elements submitted, including duplicates and invalid inputs.               |
| `summary.prohibited_numbers` | Number of entries returned in `prohibited_numbers_to_call`, including repeated inputs. |
| `summary.safe_numbers`       | Number of entries returned in `safe_numbers_to_call`, including repeated inputs.       |
| `summary.invalid_numbers`    | Number of entries returned in `invalid_numbers`.                                       |

The response always satisfies:

```text theme={null}
summary.total_checked =
  summary.prohibited_numbers +
  summary.safe_numbers +
  summary.invalid_numbers
```

Each summary count is also the length of its corresponding top-level array.

## Ordering and formatting

* Response arrays contain the original strings, not canonicalized phone numbers.
* Relative order is preserved within each of the three result arrays.
* Duplicate input occurrences remain duplicated in the output.
* Because the response is partitioned into categories, the original global
  order across all categories is not represented by a single array.

## Invalid entry

Each invalid item has this shape:

```ts theme={null}
type InvalidPhone = {
  input: string;
  reason:
    | "empty"
    | "unparseable"
    | "wrong_country"
    | "too_short"
    | "too_long";
};
```

See [Phone inputs](/guides/phone-inputs) for the complete reason-code
dictionary.

<Note>
  Earlier prelaunch examples used `matched`, `clean`, `invalid`, and `stats`.
  Those names are not part of the current public response. Update test
  integrations to use the fields documented above.
</Note>