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

# Home

> The workspace dashboard: flows over time, what needs attention, coverage by area, and recently changed documents.

**Home** is the workspace's dashboard. It is the first entry in the sidebar.

## Before the workspace is set up

A workspace that is missing either half gets two checkpoints instead of a dashboard:

1. **Connect your first context**, the documentation that says what the product promises.
2. **Connect your first repository**, the code that has to keep the promise.

Either order works. Once both are done, Home becomes the dashboard.

## Today's flows

Six cells run across the top. The first is the share of flows that succeeded, with the count under it, for example "12 of 20 flows succeeded". The other five count today's flows under the five statuses, in this order: **Succeeded**, **Failed**, **Blocked**, **Not testable**, **Never run**. Click any of them to open [Flows](/flows) narrowed to that status.

The headline counts flows, because a flow is what the engine proves and what can be proved on its own.

## Flows over time

Under the strip is a stacked chart of the same five statuses. Each point is a baseline run of one of the workspace's repositories, and the right edge is today's composition, so the chart and the strip are one picture. The period chips are **7d**, **30d**, **90d** and **All**, and it opens on 30d. Clicking a series opens Flows narrowed to that status.

Until a run has recorded its flows there is no trend, and the card says so instead of drawing an empty chart.

## Needs attention

One row per thing that is waiting on a person. There are five kinds and no more:

| Row                | What it is                                               |
| ------------------ | -------------------------------------------------------- |
| A conversation     | A piece of the agent's work that ended badly.            |
| A conflict         | Two documents that disagree and have no verdict yet.     |
| A blocked document | A document whose coverage is blocked.                    |
| A source           | A source whose last sync failed.                         |
| The provider       | The workspace has no model provider, so nothing can run. |

Each row carries its title, a status word, one fact and when it happened, and opens where that thing lives. **Agent** at the top right opens the whole list. With nothing outstanding the widget reads "Nothing is waiting on you."

## Areas, by section

One row per area of the corpus, counted **by section**, in the words the [Documents](/guard/spec-scan#documents) view uses: **Proved**, **Failed**, **Blocked**, **Not testable**, **Not run**. Each row shows a bar of the area's composition and a count per status. Clicking a row opens Documents narrowed to that area. The widget is empty until a repository reads a document.

## Recently changed, documents

The **documents** whose status a run moved, newest first, grouped into Today, Yesterday and Earlier. Each row carries the document's title, the status word it moved to, or **First read** the first time a run covered it, its reference and the time. **Failed** at the top right opens Documents narrowed to the failed ones.

## Two vocabularies, one meaning

A flow reads **Succeeded** and a section reads **Proved**. They are the same state counted in different units, and each widget names the unit it counts, so no number on Home says a different word than the page it opens.

## Next steps

<CardGroup cols={2}>
  <Card title="Flows" icon="route" href="/flows">
    The list the headline opens, and one flow's page.
  </Card>

  <Card title="Agent" icon="comments" href="/agent">
    Every piece of work, as a conversation.
  </Card>
</CardGroup>
