Sign in
latestOpenAPI 3.1.02026-08-223321,1001.6 MB

2134ebffd1ef

External API

Send a message

Send a follow-up message to an idle session.

post/v1/sessions/{session_id}/messages

Path parameters

session_idstring uuid required

Request body

inputstring required

The follow-up question or instruction. Traversal uses the full conversation history as context when responding.

thinking_mode'auto' | 'deep' | 'fast'

Investigation depth exposed on the public V1 API.

A deliberately narrower contract than the internal ThinkingMode: the lightning tier is reserved for the MCP surface and is not offered here, so unsupported values are rejected at request-parse time rather than at runtime. Values mirror ThinkingMode so the mapping stays in lockstep.

Example request

{
  "input": "Can you check if the database connection pool was exhausted during the incident?"
}

Response

Successful Response

idstring required

Unique identifier for the session.

statusstring required

Current lifecycle state of the session. One of: running (investigation in progress), idle (investigation complete, ready for follow-ups), follow_up_running (processing a follow-up message), cancelled, or failed.

titlestring nullable

Human-readable session title. Set via the title field on creation, or auto-generated by Traversal.

inputstring required

The original incident description or question that initiated this session.

created_atstring date-time required

Timestamp when the session was created, in UTC.

updated_atstring date-time nullable

Timestamp when the session was last updated, in UTC. Null if the session has not been updated since creation.

user_message_idstring required

Unique identifier for the user message that was recorded.

assistant_message_idstring required

Unique identifier for the assistant message that will contain the response. Use this ID to poll for the completed response via GET /v1/sessions/{session_id}.

Example response

{
  "id": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "status": "running",
  "messages": [
    {
      "id": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
      "confidence": "Medium",
      "confidence_explanation": "The evidence suggests a likely cause, but alternative explanations cannot be ruled out."
    }
  ],
  "user_message_id": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "assistant_message_id": "f0e1d2c3-b4a5-6789-0abc-def123456789"
}