This feature is currently available to whitelisted users only.
Go to the Chat Console > Cloud audit, select the Moderation Configuration tab, scroll to the bottom, and locate the Review result configuration module.
Review Result Configuration
You can forward review results to your app backend by setting up a basic webhook. Follow these steps:
By default, the "Return Review Result" toggle is disabled. To enable it, click Edit in the upper right corner of the Moderation Result Configuration module, turn on the return result toggle, and click Confirm.
Once enabled, review results are sent via post-webhook to the URL you specified in the basic settings. If you have not configured a basic webhook URL before enabling the return result toggle, a popup will prompt you to enter the URL. For more details, see Basic Webhook Configuration.
After activation, content callback results are delivered to your app backend through the basic webhook, supporting interception and all result configurations.
Note：
After your app backend receives a webhook request, verify that the SDKAppID parameter in the request URL matches your own SDKAppID.
API Overview
Request URL Example
In the following example, the app's webhook URL is set to https://www.example.com.
Represents the recipient of the message during recognition.
ContactType
Integer
Identifies whether the recipient is a group or C2C chat.
ContactType = 1: C2C chat.
ContactType = 2: group chat.
To_Account
String
Recipient UserID when ContactType = 1.
ToGroupId
String
Recipient GroupID when ContactType = 2.
ContentType
String
Type of recognized content:
Text = Text.
Image = Image.
Audio = Audio.
Video = Video
TextContent
String
Recognized text content when ContentType is Text.
FileURL
String
Recognized file URL when ContentType is Image, Audio, or Video.
MsgID
String
For C2C recognition, MsgID is the C2C message MsgKey; for group recognition, MsgID is the group message MsgSeq.
CtxcbResult
Integer
Recognition result from the cloud review service:
1: Intercepted
0: Allowed
CtxcbRequestId
String
Recognition identifier from the cloud review service.
CtxcbKeywords
Array
Sensitive words detected in the text.
CtxcbSuggestion
String
Recommendation from the third-party content moderation service: Block/Review/Pass.
CtxcbLabel
String
Primary label for the detected violation category: Illegal / Abuse / Polity / Ad / Porn / Sexy / Terror / Composite.
IsRestAPI
Bool
Indicates whether the webhook was triggered by active recognition:
false: triggered by internal Chat recognition
true: triggered by active recognition API, typically for audio/video recognition callbacks.
Response Example
{
"ActionStatus":"OK",
"ErrorInfo":"",
"ErrorCode":0
}
Response Fields
Field
Type
Attribute
Description
ActionStatus
String
Required
Result of request processing:
OK: Success.
FAIL: Failure.
ErrorCode
Integer
Required
Error code:
0: App backend processed successfully.
1: App backend processing failed.
ErrorInfo
String
Required
Error description.
Interception Strategy Settings
By default, only content classified as "violating" will be intercepted. If you want to intercept both suspected and violating content, or disable interception entirely, click Edit to adjust the settings. Click Confirm; changes take effect within 2 minutes.
Note：
The interception policy does not affect the moderation submission policy.
Send Interception Error Code
For synchronous image and text recognition, the interception error code is delivered by default. This means the sender receives an error code notification if their message is intercepted.
If you prefer the sender not to be notified of interception (i.e., the sender does not receive a failed delivery notification), you can disable the "Interception Error Code Delivery" toggle. To do this:
Click Edit in the upper right corner of the Moderation Result Configuration module, turn off the "Interception Error Code Delivery" toggle, and click Confirm.
Note：
To help you trace the reason for failed message delivery, you must enable the "Return Moderation Result" toggle and receive recognition results before disabling the "Interception Error Code Delivery" toggle. For details on enabling the "Return Moderation Result" toggle, see Return Moderation Result.
Violation Word Replacement Strategy
This setting determines whether violation words detected in message content are replaced. It only takes effect when interception is enabled, and will replace violation words in the message content after activation.
