Skip to main content
GET
List messages
Your messages come back as role: "user" and the agent’s as role: "assistant". Each agent reply is a single message whose content[] holds the whole reply (text, tool calls, results); a reply still in progress shows status: "streaming" with the content so far. Paging. limit (default 200, max 500) caps how much underlying detail is pulled — one agent reply is assembled from many small rows, so you get back far fewer messages than limit. When has_more is true, older detail was omitted: request the same limit with offset set to the returned next_offset to walk back through the history.

Authorizations

Authorization
string
header
required

Requires an API key. See the Authentication guide.

Query Parameters

task_id
string
required
limit
integer
default:200
Required range: 1 <= x <= 500
offset
integer
default:0
Required range: x >= 0

Response

200 - application/json

Aggregated message list. Alternates user/assistant.

data
Message · object[]
required
total
integer
required
object
string
default:list
Allowed value: "list"
has_more
boolean
default:false