Find and filter workflow runs
Open the run list, filter by status or run ID, and scope to a single workflow to find the run you need.
Every time one of your workflows fires, TaskJuice records a run. The run list under Monitoring is where you find a specific run, watch what is currently executing, and narrow a busy list down to the runs that matter. This page covers the filters that surface, what each one matches, and where to go when you need a richer view.
When to use this
Use the run list when you want to:
- Confirm a workflow actually ran after an event came in.
- Find every run that failed so you can open one and read the error.
- Watch runs that are executing right now.
- Locate one run by its ID to share or open it.
If you already know a workflow is the source and want only its runs, scope to that workflow (covered below) instead of scanning the workspace-wide list.
Open the run list
In the workspace sidebar, open the Monitoring group and select Runs. This loads the run list at /[workspaceSlug]/monitoring/workflow-runs, where [workspaceSlug] is your workspace slug. The breadcrumb reads Monitoring / Workflow Runs.
The list is a table with these columns, in order:
| Column | What it shows |
|---|---|
| Run ID | The first 8 characters of the run id, as a monospace link to the run detail. If the trigger delivered a batch, a badge next to it reads N events (or 1 event). |
| Workflow | The workflow name, linking to that workflow's own run list. Shows the first 8 characters of the workflow id when the workflow is unnamed. |
| Status | A colored badge with the run's status (for example running, completed, failed). |
| Started | The local datetime plus a relative time such as "3 minutes ago". Shows - when the run has not started. |
| Completed | Same format as Started. Shows - when the run has not finished. |
| Duration | Whole seconds with an s suffix. Shows - until both Started and Completed are set. |
The list loads the 100 most recent runs by default. To see a different run, filter the list rather than scrolling, because older runs beyond the loaded set are not shown until you narrow the query.
To pull in runs that fired since the page loaded, select Refresh in the header. The icon spins while the new data loads. The run list does not auto-refresh, so use Refresh whenever you want the latest state.
Filter the list
Two filters sit above the table. Both narrow the runs already loaded in the list.
Type a run id to find one run
Enter characters into the Run ID filter. The table keeps only rows whose run id contains what you typed. Paste the full id (or the visible 8-character prefix) to jump straight to a single run.
Pick a status to group like runs
Open the Status select and choose one value. The list keeps only runs in that status. Clear the selection to show every status again.
The Status select offers exactly five values:
| Option | Matches runs that are |
|---|---|
| Running | Currently executing. |
| Completed | Finished successfully. |
| Failed | Stopped on an error. |
| Paused | Held, awaiting a resume. |
| Cancelled | Stopped by a person. |
Both filters apply on the client, over the rows the list already fetched. They do not request more runs from the server, so they refine what is in front of you rather than searching your full run history.
Scope to one workflow
To see only the runs for a single workflow, select that workflow's name in the Workflow column. This opens its dedicated list at /[workspaceSlug]/monitoring/workflow-runs/[workflowId], with the breadcrumb Monitoring / Workflow Runs / Workflow {id}.
The scoped list is the same table, minus the Workflow column (every row is the same workflow). It adds a Run button in the header that opens a "Run workflow" dialog so you can trigger the workflow manually and watch the new run appear.
Verify you found the right run
You are looking at the run you want when:
- The Run ID column shows the prefix you filtered on, and selecting it opens the run detail.
- The Status badge matches what you expect (for example
failedfor a run you are debugging). - The Started and Completed times line up with when the triggering event arrived.
Select the Run ID link to open the full run detail, where you can read the trigger payload and step-by-step timeline. See Read a workflow run.
Troubleshooting
The run I want is not in the list. The list loads the 100 most recent runs and the two filters only narrow that loaded set. If an older run is not showing, scope to its workflow first (which loads that workflow's runs), then filter by Run ID or Status within the scoped list.
A run is stuck on running and never refreshes. The run list does not poll. Select Refresh to reload current state. If you want a view that updates on its own, open the run detail, which refreshes every 10 seconds while a run is pending or running.
There is no Paused or Waiting run in the Status filter even though a run is parked. The Status select offers five options: Running, Completed, Failed, Paused, and Cancelled. It does not list pending or waiting. A run in either of those states still appears in the unfiltered list with its real status badge; you just cannot isolate it with the Status dropdown. Filter by Run ID instead, or clear the status filter and scan the badges.
If the list shows "Failed to load workflow runs" with the message "An error occurred while loading workflow runs", the data did not load. Select Refresh to retry. If it keeps failing, confirm you have monitoring access in this workspace; the run list requires it, and a request without it is rejected before any runs are returned.
Limits
What the Monitoring run list does and does not offer, stated plainly:
- It loads up to 100 runs by default. The maximum any single load returns is 1000.
- The only filters are Run ID (free text) and Status (the five options above). There is no date-range, cost, model, or workflow-name filter, and no preset filter chips on this list.
- Filtering is client-side over the loaded runs, not a server search across all history.
- The list does not auto-refresh; use Refresh to get newer runs.
- There is no bulk delete or "clear runs" control on the standard run list.
Related
- Read a workflow run reads one run end to end: status, duration, trigger payload, and the step timeline.
- Debug a failed run walks a failed run, reads its error code and provider status, and decides whether to retry.
- Pause, resume, or cancel a run covers which control actions are allowed in each status.