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

# Scribe Sessions

> Retrieve the status and metadata of a scribe session

A scribe session is created when a clinician conducts a consultation through the Telepatia interface. Once the session is finished, you can check its lifecycle and consultation metadata here, then fetch the filled record from the [Medical Record Documents](/scribe-api/medical-record-documents) endpoint.

## Session lifecycle

A session moves through several states. The values below are returned verbatim in the `status` field.

| Status                | Phase     | Description                                                                |
| --------------------- | --------- | -------------------------------------------------------------------------- |
| `created`             | in-flight | Session opened, recording hasn't started yet                               |
| `recording`           | in-flight | Clinician is actively recording the consultation                           |
| `stopped`             | in-flight | Recording ended; audio is being uploaded and queued for processing         |
| `processing`          | in-flight | AI pipeline (transcription + record generation) running                    |
| `completed`           | **ready** | Pipeline finished — documents are available                                |
| `completedWithErrors` | **ready** | Pipeline finished with non-fatal AI errors — documents are still available |
| `reviewed`            | **ready** | Clinician has reviewed and finalized the record                            |
| `error`               | terminal  | Fatal error during processing — no documents will be produced              |
| `cancelled`           | terminal  | User cancelled the session                                                 |
| `unknown`             | —         | Fallback when the upstream state is unrecognized                           |

## Retrieving a session

Use the `consultationInternalId` you provided (or received) when setting the consultation context:

```bash theme={null}
curl https://scribe-api.telepatia.ai/v1/scribe-sessions/CONSULT-12345 \
  -H "Authorization: Bearer YOUR_API_KEY"
```

**Response:**

```json theme={null}
{
  "id": "ss_a1b2c3d4e5f6g7h8",
  "status": "completed",
  "createdAt": "2026-02-20T10:00:00Z",
  "completedAt": "2026-02-20T10:30:00Z",
  "patientName": "John Doe",
  "scribeSessionModality": "IN_PERSON",
  "consultationInternalId": "CONSULT-12345",
  "specialty": "Cardiology"
}
```

This endpoint returns lifecycle and consultation metadata only. The filled record — and its diagnosis ICD codes — is fetched from the [Medical Record Documents](/scribe-api/medical-record-documents) endpoint.

<Tip>
  While `status` is still in-flight (`created`, `recording`, `stopped`, `processing`), poll this endpoint until it reaches a ready state (`completed`, `completedWithErrors`, `reviewed`) before fetching the filled record. `error` and `cancelled` are terminal failure states — no documents will be produced.
</Tip>

## Retrieving the filled record

This endpoint returns only the session's lifecycle and consultation metadata. The actual filled record — the output of the session templates, with all sections — is fetched separately from the medical record documents endpoint, using the same `consultationInternalId`:

```bash theme={null}
curl https://scribe-api.telepatia.ai/v1/scribe-sessions/CONSULT-12345/medical-record-documents \
  -H "Authorization: Bearer YOUR_API_KEY"
```

A session can produce more than one document (one per `purpose` — e.g. `primary` for the clinician-facing record, `rpa` for EMR injection). Filter with `?purpose=primary` to fetch only the main one. See [Medical Record Documents](/scribe-api/medical-record-documents) for the full response shape and field reference.