ChatAiAgent - Realtime Compute for Apache Flink - Alibaba Cloud Documentation Center

Initiates a streaming conversation with an AI agent.

Try it now

Try this API in OpenAPI Explorer, no manual signing needed. Successful calls auto-generate SDK code matching your parameters. Download it with built-in credential security for local usage.

Test

RAM authorization

The table below describes the authorization required to call this API. You can define it in a Resource Access Management (RAM) policy. The table's columns are detailed below:

Action: The actions can be used in the Action element of RAM permission policy statements to grant permissions to perform the operation.
API: The API that you can call to perform the action.
Access level: The predefined level of access granted for each API. Valid values: create, list, get, update, and delete.
Resource type: The type of the resource that supports authorization to perform the action. It indicates if the action supports resource-level permission. The specified resource must be compatible with the action. Otherwise, the policy will be ineffective.
- For APIs with resource-level permissions, required resource types are marked with an asterisk (*). Specify the corresponding Alibaba Cloud Resource Name (ARN) in the Resource element of the policy.
- For APIs without resource-level permissions, it is shown as All Resources. Use an asterisk (*) in the Resource element of the policy.
Condition key: The condition keys defined by the service. The key allows for granular control, applying to either actions alone or actions associated with specific resources. In addition to service-specific condition keys, Alibaba Cloud provides a set of common condition keys applicable across all RAM-supported services.
Dependent action: The dependent actions required to run the action. To complete the action, the RAM user or the RAM role must have the permissions to perform all dependent actions.

Action

Access level

Resource type

Condition key

Dependent action

stream:ChatAiAgent

none

*All Resource

*

None

Request syntax

POST /advisor/v2/namespaces/{namespace}/ai-agent/stream/agent/v2/chat HTTP/1.1

Path Parameters

Parameter	Type	Required	Description	Example
namespace	string	Yes	The project name.	test-flink-bigdata-default

Request parameters

Parameter	Type	Required	Description	Example
workspace	string	Yes	The workspace ID.	710d6a64d8****
body	object	No	The resource specification.
userMessage	string	No	The natural language input from the user (1 to 64000 characters).	“”
sessionId	string	No	The session ID. If not provided, it is generated by the server. The same value must be passed for multi-turn conversations.	019F8CC7-EAD3-5E06-B0BF-3A2A0638B3DD-deliverData-20260723102220-VM8X0A5VZQ
refs	object	No	The resource references, including the job and skill lists.
jobs	array<object>	No	The list of job references.
	object	No	The task ID.
deploymentId	string	No	Deployment ID	2a63abb7-7ae7-4902-9970-fe5cff4bd7c1
jobId	string	No	Job ID	ccb853c3-1d5a-438d-bf98-346815ad875a
skills	array	No	The list of skills to inject.
	string	No	The skill.	[]
hitlDecisions	array<object>	No	The list of human-in-the-loop (HITL) approval decisions, used to resume sessions interrupted by hitlPending.
	object	No
hitlId	string	No	The ID of the approval item corresponding to the hitlPending event.	-
decision	string	No	The approval decision: approve or deny. Valid values: deny : Deny. approve : Approve.	approve

Response elements

Element	Type	Description	Example
	object
event	string	The event type.	-
sessionId	string	The ID of the current session.	462E2707-590E-51B6-9940-0AB33044828B-deliverData-202603020950-WCSN4MEC8T
text	string	The text output from the assistant.	-
completed	boolean	Indicates whether the text segment is completed (the last segment of its message).	true
toolName	string	The name of the tool function.	"get_job_events"
toolCallId	string	The tool calling ID, used to pair the invoke and result.	"c1"
displayName	string	The localized display name of the tool.	""
input	any	The key-value pairs of the tool input parameters. The structure varies based on toolName.	{ "namespace": "vvp-dev-team", "jobId": "aa91ec66-...", "deploymentId": "b78aae4c-..." }
success	string	Indicates whether the tool calling invoke is successful.	true
errorType	string	The error type on failure (only when success is false).	TIMEOUT
errorMessage	string	The error message on failure (only when success is false).	TIMEOUT
items	array<object>	The list of pending approval items.
	object	--
hitlId	string	The ID of the approval item, used when passing back hitlDecisions.	""
toolName	string	The name of the intercepted tool.	“”
args	any	The original key-value pairs of the tool parameters.	[]
displayName	string	The display name of the tool.	“”
usage	object	The token usage.
inputTokens	integer	The number of input tokens.	10
outputTokens	integer	The number of output tokens.	5
totalTokens	integer	The total number of tokens.	15
message	string	The error message for error events.	“”

Examples

Success response

JSON format

{
  "event": "-",
  "sessionId": "462E2707-590E-51B6-9940-0AB33044828B-deliverData-202603020950-WCSN4MEC8T",
  "text": "-",
  "completed": true,
  "toolName": "\"get_job_events\"",
  "toolCallId": "\"c1\"",
  "displayName": "\"\"",
  "input": "{ \"namespace\": \"vvp-dev-team\", \"jobId\": \"aa91ec66-...\", \"deploymentId\": \"b78aae4c-...\" }",
  "success": "true",
  "errorType": "TIMEOUT",
  "errorMessage": "TIMEOUT",
  "items": [
    {
      "hitlId": "\"\"",
      "toolName": "“”",
      "args": "[]",
      "displayName": "“”"
    }
  ],
  "usage": {
    "inputTokens": 10,
    "outputTokens": 5,
    "totalTokens": 15
  },
  "message": "“”"
}

Error codes

See Error Codes for a complete list.

Release notes

See Release Notes for a complete list.