Item status reporting reference
This page is the compact reference for the status cards shown above the Items list — what each count represents and how the cards relate to the full work-item status vocabulary. For the procedural walkthrough, see Item status reporting. For the statuses themselves, see Work item statuses.
The four primary cards
Most items move through four lifecycle states, and these are the cards you'll see most often. Each card shows a status and a count of items currently in that status for the AI Agent.
| Card | What contributes to the count | Underlying item status |
|---|---|---|
| In Progress | Items being processed by a flow's nodes | IN_PROGRESS |
| Review | Items paused and waiting for a person (sent to human review) | REVIEW |
| Complete | Items that finished their flow successfully (terminal) | COMPLETE |
| Failed | Items that stopped on an error before completing | FAILED |
These four carry distinct badge colors and are the states most operations teams watch.
The cards are data-driven, not a fixed set of four
The cards are driven by your data, not hard-coded to four. A card appears for each status that currently has at least one item. So alongside the four above, you may also see cards for Queued (QUEUED), Deferred (DEFERRED), and Closed (CLOSED) when items are in those states. Items in the Canceled (CANCELED) state are never shown as a card.
This is why the mapping is not one-to-one with a four-status model: the four are simply the most frequently present cards, while the full item vocabulary has eight statuses. See Work item statuses for the complete set.
What the counts represent
- Scope: counts are for the current AI Agent.
- Whole filtered set: a count reflects every matching item, not just the page on screen, so a count can exceed the number of visible rows.
- Follow your filters: when you filter the list (by status, reviewer, flow, date, or search), the cards recount to match the same filtered set, so the cards and the list always agree.
- Point-in-time: counts refresh when the list loads; reload to bring them up to date. They are a count as of the last load, not a live ticker.
Not an analytics dashboard
The cards are a quick operational read — current counts only. They do not provide trends, throughput, straight-through rates, or any historical analytics. The same item statuses are exposed through the API; the cards simply summarize their counts for the current AI Agent and filters.
Drilling in
The cards summarize counts; to act on items, use the list below them. Filter by status to narrow to one state (the status filter offers every state, including those not shown as a card), then open an item to see its Insights and Review detail. To read the error behind a failed item, an Admin or Partner opens the run in Run history.
Where to go next
- Item status reporting — the full walkthrough of the cards and list.
- Work item statuses — the complete eight-status vocabulary.
- Human review — act on items in the Review status.
- Items — item statuses in the API.