Skip to main content
GET
Get an agent run

Authorizations

X-API-Key
string
header
required

Path Parameters

agent_id
string<uuid>
required

The ID of the agent that owns the run.

run_id
string<uuid>
required

The ID of the run to retrieve.

Query Parameters

verbose
boolean
default:false

Include each step's raw outputs payload in the execution trace.

Response

Successful Response

Status and result details for an agent run.

id
string<uuid>
required

Unique ID for the run.

agent_id
string<uuid>
required

Unique ID of the agent for this run.

status
enum<string>
required

Current status of the run.

Available options:
queued,
running,
succeeded,
failed,
cancelled,
skipped,
unknown
started_at
string<date-time> | null

When the run started, if it has started.

finished_at
string<date-time> | null

When the run finished, if it has completed.

error
Error · object | null

Error details, when the run fails and error information is available.

outputs
Outputs · object

Output values returned by the run, keyed by output-variable UUID. This UUID-keyed object is retained for compatibility and is empty when no outputs are available.

outputs_expanded
Outputs Expanded · object | null

Expanded form of outputs, keyed by the same output-variable UUIDs. Each entry carries the agent's configured human-readable key as title alongside the returned value. title is null when the agent has no configured key for that output. Entries preserve the key order of outputs. The UUID-keyed outputs field remains the stable compatibility field.

steps
AgentRunStep · object[] | null

Ordered step-by-step execution trace — one entry per node that ran, in execution order. Always present once the run has executed a node; per-node outputs inside each step are included only when the request asks for verbose.