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

# Get a run

> Poll a launched run and read the finished audit.

```http theme={null}
POST https://api.humalike.com/v1/social-observability/projections/audit-run
```

Read a run: its stored transcript and every section of the audit that has
finished so far. This is how you collect the result of
[`audit_launch`](/api-reference/audit/launch).

The call is free and read-only. Poll it every few seconds after launching.

## Authorization

<ParamField header="Authorization" type="string" required>
  Your bearer token: `Bearer <token>`. See [Authentication](/authentication).
</ParamField>

## Request body

<ParamField body="run_id" type="string" required>
  The run to read.
</ParamField>

## Request

```bash theme={null}
curl https://api.humalike.com/v1/social-observability/projections/audit-run \
  -H "Authorization: Bearer $HUMALIKE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "run_id": "9f8b1c04-6d2a-4f31-9c77-2b1d0e5a8c43" }'
```

## Knowing when it is done

Sections appear as they finish, so an early poll returns a partly filled
response. The audit is finished when **`report`, `read`, and `verdicts` are all
non-null**:

```python theme={null}
import time, httpx

while True:
    run = httpx.post(url, headers=headers, json={"run_id": run_id}).json()
    if run["report"] and run["read"] and run["verdicts"] is not None:
        break
    time.sleep(5)

print(run["report"]["health_score"], run["report"]["summary"])
```

`replies` fills in shortly after those three, so read it last. A run typically
completes in a few minutes; if all three are still null after 30 minutes, treat
the run as failed and prepare a new one.

## Response

<ResponseField name="run_id" type="string">
  The run's id.
</ResponseField>

<ResponseField name="agent_name" type="string | null">
  The agent under review. `null` on a run that was prepared but never launched.
</ResponseField>

<ResponseField name="transcript" type="object">
  The parsed transcript the audit ran on, as `{ "messages": [...] }`. Each
  message carries the `id` (`m1`, `m2`, … in order), `speaker`, and `text`. The
  ids are what the report's evidence cites.
</ResponseField>

<ResponseField name="report" type="object | null">
  The reception report: `health_score`, `summary`, `findings`, `per_user`,
  `interactions`, and `interaction_totals`. Same shape as
  [`analyze`](/api-reference/analyze). `null` until that section finishes.
</ResponseField>

<ResponseField name="read" type="object | null">
  What your agent could have known about the people in the conversation.

  <Expandable title="read">
    <ResponseField name="prompt_block" type="string | null">
      The context, rendered as a block you can paste into a system prompt.
    </ResponseField>

    <ResponseField name="portrait" type="object | null">
      A portrait of your agent as the transcript presents it.
    </ResponseField>

    <ResponseField name="mental_state" type="array | null">
      Per person: their inferred `beliefs`, `goals`, and `emotions`.
    </ResponseField>

    <ResponseField name="profiles" type="array | null">
      Per person: the `facts` a memory would have retained about them.
    </ResponseField>
  </Expandable>

  Every field inside is independently optional — a section that could not be
  produced is omitted rather than failing the audit.
</ResponseField>

<ResponseField name="verdicts" type="array | null">
  One entry per turn your agent took, each with the `index` of the turn's first
  message, a `risk` grade, a `summary` of what the turn risked, and
  `predicted_message` — what the other person was likely to say next.
</ResponseField>

<ResponseField name="replies" type="array">
  Rewrites of your agent's worst turns. Each has the `index` of the turn, the
  rewritten `reply`, `messages` (the reply split the way a person would send
  it), and the `risk` grade of the rewrite. Empty until the rewrites finish.
</ResponseField>

## Errors

| Status | Code               | Cause                                              |
| ------ | ------------------ | -------------------------------------------------- |
| `400`  | `VALIDATION_ERROR` | The run does not exist, or does not belong to you. |
| `401`  | `UNAUTHORIZED`     | Missing or invalid bearer token.                   |

Next: [Analyze a transcript directly](/api-reference/analyze) when you already
have structured messages and want just the report.