> ## Documentation Index
> Fetch the complete documentation index at: https://docs.unify.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitoring tasks

> The Tasks tab — every definition and every run, in one place

The **Tasks** tab lives in your assistant's workspace rail, and it exists to
answer two questions at any moment: *what standing work does this teammate
have*, and *how has it been going?* In the tab's own words, it holds
"scheduled, recurring, triggered, and continuous workflows — definition and
run history together."

That "together" is the design: there's no separate activity page to
cross-reference. Each task carries its own history, so the setup and the
track record are always one click apart. (The ⓘ button in the tab header
opens a quick "How to use this tab" guide if you ever want the refresher
in-product.)

## The task list

Tasks appear as expandable cards. Collapsed, each shows the essentials on
one line: the task's name, its cadence in plain words ("Every week on Mon
at 08:00", "Every 30 minutes"), when it next runs, its priority, and a
status badge. The footer keeps the running tally — how many tasks exist and
how many runs have been logged.

Use the **All / Active / Paused** filter and the search box to narrow
things down, and the refresh button to pull the latest. When any task is
mid-run, a pulsing **Working** indicator appears in the toolbar.

Expand a card and it splits in two:

* **The definition** (left) — the plain-language description your assistant
  works from, plus the structured facts: **Type** (Scheduled, Triggered,
  On demand, Offline), **Trigger** (which event arms it, for triggered
  tasks), **Cadence**, **Start**, **Next run**, and **Priority**.
* **Run history** (right) — every past run in a compact table: its state,
  why it started ("On schedule", "Triggered by Email", "Started on
  demand"), when it started and finished, and how long it took.

Click any run to open the **Run Detail** view — "what happened, why it
started, and when it ran": the task it belongs to, its status, the contact
and channel that triggered it where relevant, and the precise timing from
scheduled-for through completed-at.

## What the statuses mean

| Badge         | Meaning                                               |
| ------------- | ----------------------------------------------------- |
| **Scheduled** | Will start automatically at its next scheduled time   |
| **Ready**     | Armed and waiting for a matching event                |
| **Running**   | Actively executing right now                          |
| **Completed** | Finished successfully                                 |
| **Failed**    | Stopped because something went wrong during execution |
| **Cancelled** | Was stopped before it finished                        |
| **Pending**   | Created and waiting to start                          |

## What to use the tab for

In practice, the Tasks tab earns its place in a few recurring moments:

* **Right after creating a task** — confirm it landed the way you meant:
  the description reads right, the cadence parsed correctly, the next run
  is when you expected.
* **The "did it run?" check** — a glance at run history answers whether
  last night's job fired, how long it took, and whether it succeeded.
* **Triaging a failure** — a **Failed** badge plus the run's detail view
  tells you when and how a run died, so you can ask your assistant a
  precise question about it.
* **Auditing the standing workload** — a periodic scroll through the list
  keeps the set of recurring jobs intentional rather than accreted; anything
  stale is one chat message away from being cancelled.

## Watching a run live

Run history tells you what happened; the
[**Actions** tab](/learning/watching-it-work) shows what's happening. When a
task fires, its work streams there in real time — handy the first time a new
task runs, or when you've hit **Test it** on a trigger and want to watch the
dominoes fall.

## Making changes

The Tasks tab is a window, not a control panel — creating, editing, pausing,
and deleting all happen [by asking your assistant](/tasks/creating). If
something in the list looks off, the fastest fix is to say so: *"the Friday
recap ran twice last week — what happened?"*