Message the "Revise Data Product" Agent (streaming)

post

{protocol}://{base-url}/ai/api/v1/chats/agent/default/revise_data_product_agent/stream

Post a new Message to the Chat using the "Revise Data Product" Agent.

Query Params

chat_id

The ID of the Chat to use. If not provided, a new Chat will be created.

tags

array | null

If creating a new chat, add these tags to it. This is ignored if the chat already exists.

Body Params

The payload to be sent to the agent. This should conform to the input JSON schema of the agent. Optionally, a 'tool_approvals' parameter may be accepted with a dictionary of tool_call_id to approval boolean. When submitting tool approvals, the input payload must match the original inputs used to generate the tool calls (same as the previous API call). If tool_approvals is provided, the message parameter is ignored and the tool calls are executed or skipped.

tool_approvals

object | null

[LEGACY] A mapping of tool_call_id to approval boolean. When submitting tool approvals, the input payload must match the original inputs used to generate the tool calls (same as the previous API call). If tool_approvals is provided, the message parameter is ignored and the tool calls are executed or skipped. Prefer using 'approvals' field for new integrations.

approvals

object | null

A mapping of function name to approval decision for codemode agents. Applies to functions decorated with @requires_approval. Use 'approve' for one-time approval, 'approve_session' for session-level approval, 'deny' for one-time denial, or 'deny_session' to deny for the session.

message

string

required

length ≥ 1

The message to send to the agent.

data_product_id

string

required

length ≥ 1

The ID of the data product to evaluate.

data_product_version

The version of the data product to evaluate. If None, uses the default version.

central_eval_agent_config_id

Agent config ID to use for central SQL evaluations.

central_eval_case_ids

array | null

Evaluation case IDs for central SQL evaluations.

central_eval_input_payload

object | null

Input payload for the central SQL evaluation agent.

central_eval_custom_llm_judge_prompt

Custom LLM judge prompt for central SQL evaluations.

revise_dp_result_id

Revise data product result ID, used by the central SQL evaluation tool to enable
chat resumption after evaluation completes.

Headers

string

enum

Defaults to application/json

Generated from available response content types

Allowed:

Responses

204

The message is empty and was ignored.

Language

Credentials

URL

Loading…

Response

Choose an example:

application/json

text/event-stream

201Message sent successfully. Returns an event stream of Message objects.

401Authentication required. No valid session cookie or bearer token provided, or authentication failed.

402Tenant has no remaining tool calls. The request cannot be completed.

404The chat could not be found.

409Chat is busy. Returns an event stream of the running chat.

413The payload exceeds the model's context window.

422The payload is not valid according to the input JSON schema of the agent.

500Internal server error during request processing.