Process text or preview a saved assistant draft

The preview_assistant_draft operation is available in every workspace and requires a Bearer session, selected X-Workspace-Id membership, and draft-editor or administrator access. Other operation contracts are unchanged. Preview uses a saved current draft, a reviewed safety baseline, current knowledge and model, and no tools or persistent conversation.

POST

https://api.channelautomation.com/api/ai/process-text

Request body JSON required

workspaceId
string
required

operation
string
required
text
optional
string
options
optional
object
- agentId
  optional
  string
  format: uuid
- draftToken
  optional
  integer
  minimum: 1
- messages
  optional
  array of
  object
  role
  optional
  user
  or
  assistant
  content
  optional
  string
  maxLength: 8000
- simulatedContact
  optional
  object
  Synthetic display fields only; real contact IDs are ignored.
}

Returns

200

Generated text; preview also returns model and draftToken

400

Invalid input

401

Preview session required

403

Preview access denied

409

Saved draft changed

503

Preview could not run

Process text or preview a saved assistant draft

curl --request POST \
  --url https://api.channelautomation.com/api/ai/process-text \
  --header 'Content-Type: application/json' \
  --data '
{
  "workspaceId": "string",
  "operation": "string",
  "text": "string",
  "options": {
    "agentId": "bc309ecf-5f66-4057-93c5-6611cc9cb7b2",
    "draftToken": 1,
    "messages": [
      {
        "role": "user",
        "content": "string"
      }
    ],
    "simulatedContact": {}
  }
}
'

const url = 'https://api.channelautomation.com/api/ai/process-text';
const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    workspaceId: 'string',
    operation: 'string',
    text: 'string',
    options: {
      agentId: 'bc309ecf-5f66-4057-93c5-6611cc9cb7b2',
      draftToken: 1,
      messages: [{role: 'user', content: 'string'}],
      simulatedContact: {}
    }
  })
};

fetch(url, options)
  .then(res => res.json())
  .then(json => console.log(json))
  .catch(err => console.error(err));

import requests

url = "https://api.channelautomation.com/api/ai/process-text"

payload = {
    "workspaceId": "string",
    "operation": "string",
    "text": "string",
    "options": {
        "agentId": "bc309ecf-5f66-4057-93c5-6611cc9cb7b2",
        "draftToken": 1,
        "messages": [
            {
                "role": "user",
                "content": "string"
            }
        ],
        "simulatedContact": {}
    }
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)