Read one step of a session's trace in full: tool input and output, the model request and response, and its raw events.
const url = 'https://app.everruns.com/api/v1/sessions/example/trace/turns/example/steps/example';const options = {method: 'GET'};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request GET \ --url https://app.everruns.com/api/v1/sessions/example/trace/turns/example/steps/exampleParameters
Section titled “ Parameters ”Path Parameters
Section titled “Path Parameters”Prefixed public identifier
Prefixed public identifier
Prefixed public identifier
Query Parameters
Section titled “Query Parameters”Return payloads whole, however large.
Responses
Section titled “ Responses ”Success
Everything about one step, for the inspector.
object
Raw events of the step, in sequence order.
An event in the session log.
object
Payload, when small enough to inline.
Event identifier.
Sequence in the session event log.
Size of the serialized payload.
When the event happened.
Event type, such as tool.call_completed.
The request of a model call: what changed since the previous call.
object
Messages sent in the call.
Messages before this index were already in the previous call.
The new messages, at most 50.
One message of a model request.
object
The whole message, when asked for and within the limit.
Position in the request.
Start of the message text.
system, user, assistant or tool.
Size of the serialized message.
Start of the system prompt.
Tools offered in the call.
The step itself.
object
Session a spawn_agent call started.
Wall time of the step; absent while running.
Last event sequence of the step.
Prompt tokens of a model call.
model, answer, tool, approval, agent or send.
Model that served the call, for model steps.
Tool name, or model for model calls.
What the model said alongside this call (model calls and answers).
Start relative to the turn’s start.
Completion tokens of a model call.
Tool calls this model call asked for.
Short result or error.
First event sequence of the step.
When the step started.
running, success, error or cancelled.
Per-turn step number, 1-based.
What the call acted on, from the tool’s narration.
Tool call identifier, for tool steps.
Turn the step belongs to.
Example
{ "events": [ { "id": "event_01a124cd8a1b7f02a5c4e6d1b2f3a4c5", "sequence": 14, "size_bytes": 512, "type": "tool.started" } ], "input": { "size_bytes": 221, "truncated": false }, "output": { "size_bytes": 221, "truncated": false }, "request": { "message_count": 14, "new_from": 12, "new_messages": [ { "index": 12, "role": "tool", "size_bytes": 244 } ], "tool_count": 9 }, "step": { "duration_ms": 269, "end_sequence": 21, "kind": "tool", "model": "gpt-6.1-sol", "name": "bash", "offset_ms": 3010, "start_sequence": 14, "status": "success", "step": 2, "tool_call_id": "call_PzDcvnFaUll8os3tPhWj", "turn": 1240 }}