---
title: "Draft an answer with AI"
method: POST
path: "/room-questions/{question_id}/draft"
tags: ["dataroom"]
---

# Draft an answer with AI

`POST /room-questions/{question_id}/draft`

## Request

- Base URL: `https://localhost:4410/api/v1`
- URL: `https://localhost:4410/api/v1/room-questions/{question_id}/draft`
- Auth: HTTP bearer

A draft answer to a reader's question on the workspace's own language model key, from pages of the documents the asker's link opens, each fact cited by page. Citations to other pages are taken out, and a draft that cites none isn't shown. Sentences about future performance are flagged by words and, with a decision model, by its answer. Off unless an owner turned on drafting answers with AI. Nothing is published: dataroom.question.answer publishes a member's text.

## Path parameters

- `question_id` string, required

## Request body

- object

## Response `200`

The result.

- object
  - `answer` string, required
  - `model` string, required
  - `citations` object[], required
    - `label` string, required
    - `document_id` string, required
    - `page` integer, required
  - `left_out` integer, required
  - `flags` object[], required
    - `sentence` string, required
    - `source` 'rule' | 'model', required
    - `band` 'high' | 'medium', nullable, required

## Other responses

- `default`: An RFC 9457 problem.

---

[API](https://skmtc.dev/binderwork/apis/binderwork-api.md) · [All operations](https://skmtc.dev/binderwork/apis/binderwork-api/llms.txt) · [OpenAPI document](https://skmtc.dev/binderwork/apis/binderwork-api/revisions/f8fab0ebee53?raw)