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

# Analytics

> See how your projects, workflows, and teams are actually using MagOneAI - executions, tokens, activities, and errors

## What analytics gives you

Analytics turns the raw activity of your workflows into dashboards you can read at a glance. Every dashboard aggregates what has already happened - executions, token usage, activity performance, and errors - scoped to a **project**, a single **use case**, or a **workspace (organization)**.

<CardGroup cols={3}>
  <Card title="Project" icon="folder">
    A rollup across every workflow in a project: execution health, token trends, activity mix, and the top workflows and users.
  </Card>

  <Card title="Use case" icon="diagram-project">
    A single workflow in detail: how it was triggered, how each activity performed, its most common errors, and who consumed tokens.
  </Card>

  <Card title="Workspace" icon="building">
    An organization-wide view: counts, token usage, execution status, top projects, and a breakdown by model.
  </Card>
</CardGroup>

## Choosing a time window

Every dashboard takes a `days` window (default 7, up to 90). Set it to **0** to see **all-time** analytics - the window is floored at the beginning of your data rather than a fixed number of days back.

<Note>
  On an all-time window there is no prior period to compare against, so the period-over-period change indicator is simply omitted.
</Note>

## Project analytics

The project dashboard answers "how is this project doing overall?"

* **Execution summary** - total runs, how many completed, failed, are still running or pending, the success rate, and average duration, with a change indicator versus the previous period.
* **Execution and token trends** - daily time series so you can see volume and consumption rise and fall.
* **Activity breakdown** - counts and average duration by activity type (Agent, Tool, and so on), so you can see where the work concentrates.
* **Top workflows** - the use cases with the most runs.
* **Top users** - the most active members, shown by **display name** rather than email.

You can also scope the whole dashboard to **just your own** executions.

## Use case analytics

The use-case dashboard drills into one workflow:

* **Execution summary and trend** - the same health stats, for this workflow alone.
* **Trigger breakdown** - how runs were started: In-App, Chat, Schedule, or API.
* **Activity performance** - per-activity metrics (run count, successes, failures, success rate, average duration), with each activity resolved to its name in the workflow.
* **Top errors** - the most common failures, **sanitized** (file paths stripped, message truncated) so dashboards never leak internals.

### Token analytics: input vs output

Token usage is tracked as a daily time series that splits **input tokens** from **output tokens** (plus the total), so you can tell reading-heavy workloads from generation-heavy ones. Alongside it, **per-user token usage** shows who consumed how many tokens in this use case - total tokens, execution count, and last-active time - again by display name only, ordered by tokens descending.

<Tip>
  Token analytics show consumption; [Usage & quotas](/platform/usage-and-quotas) is where you turn that consumption into enforced limits. Use the two together: watch here, cap there.
</Tip>

## Workspace analytics

At the organization level, admins and owners see a full workspace dashboard - entity counts, token usage, execution status breakdown, top projects, and token usage grouped by model configuration and by source. Members of the same workspace see the same shape scoped to **their own** activity. Superadmins can additionally view a specific user's activity within a workspace.

## Excel export

Analytics can be exported to an **Excel workbook** for offline reporting or sharing:

* **Workspace export** - comprehensive per-project analytics across the whole organization. Available to superadmins, org admins, and org owners.
* **Per-user workspace export** - one user's activity within a workspace. Available to superadmins.

Exports default to a 30-day window (still up to 90, or all-time with `days=0`) and download as a dated `.xlsx` file.

## Privacy and access

* Every dashboard checks authorization before returning anything. The use-case dashboard returns a **not-found** response for a caller without access, so it never hints that the workflow exists; the project and workspace dashboards return an **access-denied** error instead.
* People are shown by **display name**, never email, in every ranking and per-user view.
* Error messages are sanitized before they appear.

## Platform-wide stats are separate

The dashboards above are for **your** projects, workflows, and workspace. A distinct, **superadmin-only** platform dashboard reports metrics across the entire deployment - totals for organizations, projects, agents, and executions, platform-wide token usage, and top organizations. That view lives in the Admin Portal and is not part of the per-workspace analytics described here.

<CardGroup cols={2}>
  <Card title="Admin Portal" icon="gear" href="/platform/admin-portal">
    Where superadmins see platform-wide statistics across every organization
  </Card>

  <Card title="Usage & quotas" icon="gauge" href="/platform/usage-and-quotas">
    Turn token consumption into enforced limits
  </Card>
</CardGroup>

## Next steps

<CardGroup cols={2}>
  <Card title="Triggers & execution" icon="play" href="/workflows/triggers-and-execution">
    Where each run - and its token usage - is recorded
  </Card>

  <Card title="Workflows overview" icon="diagram-project" href="/workflows/overview">
    The use cases these dashboards measure
  </Card>
</CardGroup>