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

# Project overview

> The project Overview shows spend or tokens for a period, a live burn rate, and breakdowns by model, session, person, source, and pull request.

Every project opens on its **Overview**: one page that answers how much your agents used, who used it, on which models and tools, and what it produced. Open it from the **Projects** home by selecting a project, or from **Overview** in the left rail.

## Navigating the dashboard

* **Left rail:** **Projects** (your organization's home), then the project's pages — **Overview**, **Sessions**, **Users**, **Skills**, **MCP**, **Burn map**, **Pull requests**, **Metrics** (**Dashboards**, **Explore**), and **Models** (**Usage**, **Compare**). **Settings** sits at the bottom.
* **Top bar:** switch organization or project, create a new one, and see the live status of the page's data (**Ready**, **Updated**, **Refreshing…**).

## Period and cost view

Two controls in the page header apply across the dashboard, and your choice is remembered:

* **Period:** `24h`, `7d`, or `30d` (the default).
* **Usage view:** **Tokens**, **Notional** (the default), or **Net**. See [Cost views](/concepts/entitlements#cost-views) for what each one counts.

## Headline and live burn rate

The top card shows **Spend** (or **Tokens**) for the period, with its trend against the period before, and how much of it a subscription covered versus metered billing. Below it are the period's **Tokens**, **Sessions** (with turns), **Active users** (with API keys), and **PRs merged**, plus a token activity calendar.

**Burn rate now** is live: tokens per minute and dollars per hour over the last 60 seconds, the number of calls in that minute, a last-hour sparkline by token type, and **Burn by model** and **Burn by user** rankings.

## Breakdowns

Below the headline, the period is compared with the one before it:

| Panel | What it shows |
| - | - |
| **Spend by hour** / **by day** | Spend or tokens over the period, stacked by model |
| **Model mix** | Each model's share of tokens against its share of spend; select one to open its [model page](/dashboard/models) |
| **Sessions** | The top sessions, by cost or by time |
| **People** | The people using the project, by spend |
| **Spend by source** | Spend per source (`claude-code`, `codex`, `openrouter`, …) with its session count |
| **Recent pull requests** | Pull requests touched by agent sessions in the last 7 days |

Select a source in **Spend by source** to open that source's page: its tokens, sessions, active users, and model calls, spend over time, and its models, providers, principals, and sessions.

## Banners

The Overview tells you when something needs attention:

* **Funding gap** — people are running an agent without plan usage being captured, so Terma cannot tell what a subscription covered. The fix is to install the Terma CLI for them.
* **Unmapped principals** — identities that are not mapped to a member of your organization. See [Principal mapping](/dashboard/settings#principal-mapping).
* **Connect GitHub** — pull request data needs the GitHub connection. See [Pull requests](/dashboard/pull-requests#connect-github).

## From the CLI

The same numbers are available in a terminal:

```bash theme={null}
terma usage --since 7d                    # spend by person
terma usage --since 7d --group-by source  # spend by source
terma usage --since 7d --group-by model   # spend by model
```

<Note>
  `terma usage` reports provider cost, which corresponds to the **Notional** view.
</Note>