> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.simplified.com/mcp/jobs-and-results/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.simplified.com/_mcp/server.

# Jobs and results

Generation, media processing, asset ingestion, and workflow execution can outlast a single client request. Some connector tools wait for completion automatically, but you should not assume every successful call contains a finished output.

## Read the response before continuing

Check for a completed output, a processing state, an error, and any identifiers. Keep the original response with its teamspace so another call can retrieve the same work.

| Operation                           | Follow-up                                                         |
| ----------------------------------- | ----------------------------------------------------------------- |
| Task-based generation or processing | `api_getTaskResult` with the returned task ID                     |
| Video generation                    | `api_getVideoVariation` with the returned video and variation IDs |
| Transcription                       | `media_getTranscription` with the transcription ID                |
| Asset import or upload              | `api_getAsset` with the asset ID                                  |
| Workflow execution                  | `flows_getWorkflowRunStatus`, then `flows_getWorkflowRun`         |

Use the follow-up that matches the initial response. An asset ID, transcription ID, and task ID identify different resources.

## Poll an existing job

Wait between status checks and respect any retry guidance. Stop on a documented success or failure state. Keep passing the same `space_id` when you make follow-up calls yourself.

> Check the existing video generation job. If it is still running, keep its identifiers and report its status. Do not start another generation.

## Recover from a timeout

A client timeout does not establish that the server canceled the job. Retrieve the job or created resource before resubmitting. Repeating a generation request can consume more credits; repeating a create request can produce a second item.

If no identifier was returned, inspect relevant recent resources where possible. Report the uncertainty instead of claiming failure or silently recreating the work.

## Report a useful result

Return the actual status, saved IDs, and available output links. On failure, include the error and the next action needed. Do not describe a submitted request or a queued execution as completed content.

> Track long-running work without creating duplicate jobs.