Sign in

v2

latestOpenAPI 3.1.02026-08-05267431678.1 KB
auto-agent

Get Run Status

Get the status and results of an Auto Agent run.

Poll this endpoint to check if the run is still running or has completed. When the run is active and new logs exist, schedules a background task to generate an AI progress summary.

Args: run_id: UUID of the run to check. auth: Authenticated user (ownership is enforced).

Returns: Current run status and results if complete.

Raises: HTTPException: 404 if run not found or not owned by user.

get/auto-agent/run/{run_id}

Path parameters

run_idstring required

Response

Successful Response

run_idstring required

Run UUID

statusstring required

running | complete | failed | stopped

messagestring

User task description

project_idstring nullable

Project ID grouping datasets and models from this run

deliverables_jsonobject nullable

Parsed deliverables.json (available when complete)

curation_reportstring nullable

data-curation.md contents (available when complete)

final_reportstring nullable

final_report.md contents (available when complete)

error_messagestring nullable

Error description (available when failed)

tool_calls_madeinteger

Number of tool calls executed

created_atstring required

ISO timestamp of run creation

completed_atstring nullable

ISO timestamp of completion

current_phasestring nullable

High-level progress phase: analyzing | researching | preparing_data | training | evaluating | finalizing

progress_updatesobject[] nullable

Accumulating AI-generated progress summaries, each with summary and timestamp