Continue Conversation

When Start Conversation returns status = interrupted, the Agent is waiting for more information from you. Send your answers via this endpoint and the paused run resumes from the interrupt point, until it succeeds, gets interrupted again, or fails.

A single round of conversation may be interrupted multiple times: if the run returns interrupted again after continuing, call this endpoint again with the new interrupt.

As with Start Conversation, choose blocking / SSE streaming response mode with the Accept header.

SDK Links

Python	longbridge.openapi.AgentContext.continue
Rust	longbridge::agent::AgentContext#continue
Go	AgentContext.Continue
Node.js	AgentContext#continue
Java	AgentContext.getContinue
C++	longbridge::agent::AgentContext::continue

Request

HTTP Method	POST
HTTP URL	/v1/ai/agents/:id/conversations/:chat_uid/messages/:message_id/continue

Path Parameters

Name	Type	Required	Description
id	string	YES	UID of the target Agent
chat_uid	string	YES	chat_uid of the interrupted conversation, from the Start Conversation response
message_id	string	YES	ID of the interrupted message, from the message_id in the interrupt response

Request Body

Name	Type	Required	Description
answers_by_tool_call	object	YES	Answers keyed by tool_call_id; must not be empty. Each value is a map of question text → answer

The structure of answers_by_tool_call is "tool call ID → answers for that inquiry", where each answer set is a map of "question → answer":

• The outer key corresponds to tool_call_id in the interrupt structure.
• The inner key is the interrupt question text, and the value is the answer you selected or typed.
• All questions required by the interrupt must be answered, otherwise an error is returned.

Request Example

curl -X POST "https://openapi.longbridge.com/v1/ai/agents/123/conversations/ct_9f2c1a5b/messages/43/continue" \
  -H "Authorization: Bearer <token>" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json" \
  -d '{
    "answers_by_tool_call": {
      "call_abc123": {
        "Which time range would you like to check?": "Past month"
      }
    }
  }'

curl -N -X POST "https://openapi.longbridge.com/v1/ai/agents/123/conversations/ct_9f2c1a5b/messages/43/continue" \
  -H "Authorization: Bearer <token>" \
  -H "Content-Type: application/json" \
  -H "Accept: text/event-stream" \
  -d '{
    "answers_by_tool_call": {
      "call_abc123": {
        "Which time range would you like to check?": "Past month"
      }
    }
  }'

Response

The response structure is identical to Start Conversation.

Response Headers

• Content-Type: application/json (blocking)
• Content-Type: text/event-stream (streaming)

Response Example

Run succeeded:

{
  "code": 0,
  "message": "success",
  "data": {
    "chat_uid": "ct_9f2c1a5b",
    "message_id": "43",
    "status": "succeeded",
    "answer": "Over the past month, Tesla (TSLA.US)...",
    "references": [],
    "elapsed_time": 2.74
  }
}

Response Status

Status	Description	Schema
200	Success	conversation_response
400	Invalid parameters	Empty answers_by_tool_call, message not interrupted, missing required interrupt answers, or conversation / message ownership mismatch
500	Internal error	None

Ownership Checks

To prevent resuming someone else's run, the server verifies that the conversation identified by chat_uid belongs to the authenticated member, and the message identified by message_id belongs to that conversation. Failing either check returns 400.