> For the complete documentation index, see [llms.txt](https://docs.rootcause.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.rootcause.ai/user-guide/reports.md).

# Produce Reports

Simulations produce findings. Reports package those findings into a document you can share — with an executive, a client, or a colleague who doesn't work in the platform.

RootCause reports are living documents. Charts and metrics stay connected to live data, so a report you build today reflects updated numbers tomorrow without manual edits. They also serve as an audit trail: by linking conclusions to specific Datasets, Digital Twins, and simulation runs, you document not just what you found but how you found it.

***

## Create a report

Open **Reports** in the top navigation bar and click **New report**.

<figure><img src="https://1662811113-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FBXg3gZLR0e2Q8SzeQmql%2Fuploads%2Fgit-blob-b5ff590625c542725681fffc8ec3770fe7400597%2Frc-reports-home.png?alt=media" alt="Reports home with the New report button and one report listed with a Ready status"><figcaption><p>The Reports home lists every report in the workspace with its owner and status. <strong>New report</strong> starts a draft.</p></figcaption></figure>

**New report** creates a draft and opens the blank editor directly. You then write it yourself, or have the agent write it from inside the editor: press `/` and choose **Ask agent…**, or select some text, right-click and choose **Edit with AI**.

***

## The report editor

<figure><img src="https://1662811113-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FBXg3gZLR0e2Q8SzeQmql%2Fuploads%2Fgit-blob-9a451cfff8f7308effed758e57e913e6082d5d41%2Frc-report-editor.png?alt=media" alt="The report editor showing a generated report titled Causal Drivers of Customer Churn and the Effectiveness of Service Interventions, with a formatting toolbar, Insert widget, Share, Export PDF and Save controls, a Contents outline on the right, and an Executive Summary whose key numbers are evidence links"><figcaption><p>A generated report in the editor. Sections are structured, the Contents rail tracks the outline, and the highlighted numbers are evidence links back to the underlying runs. The report can be edited freely after generation.</p></figcaption></figure>

The editor works like a modern document tool. Type to add text; headings, bullet and numbered lists and block quotes come from the formatting toolbar, or from markdown shortcuts as you type (`#` , `-` , `>` ).

`/` opens the **Insert Widget** palette, which carries **Ask agent…**, the report widgets (KPI Card, Stat Grid, Chart, Table, Source Preview, Dataset, Causal Graph, Digital Twin, Markdown, Heading, Divider, Image, Iframe) and **Evidence Reference**. It filters as you type, so `/chart` or `/kpi` finds the widget you want.

**Widgets** connect directly to Datasets and simulation results. A chart widget referencing a Dataset updates automatically when the underlying data changes — no manual refresh or re-export needed.

**References** — workspace objects the agent cites appear inline as chips linking to the object, so readers can follow them to inspect the underlying analysis. Typing `@` reaches the object picker in the Assistant's chat input, not in the report body; in the report, add citations with the **Evidence Reference** command.

***

## Evidence and audit trail

For conclusions that will drive decisions, link them to evidence:

1. Place the cursor where the citation should go
2. Press `/` and choose **Evidence Reference**, or use the command palette's **Insert Evidence**
3. In the **Insert Evidence Reference** dialog, pick a record from the evidence already attached to this report
4. Optionally set a **Display Value** to override what is shown, then click **Insert Reference**

The list is populated by evidence the agent gathered while working on this report, so a hand-written report with no agent activity shows "No evidence available in this report".

The evidence link is embedded in the report. Readers can click through to verify the data behind each conclusion. This is what separates a report from a presentation: the reasoning is inspectable, not just asserted.

***

## Export and share

**PDF Export** — click **Export PDF** in the top right to generate a polished PDF snapshot. Suitable for sharing outside the platform with clients or leadership. The PDF is static — it captures the report at the moment of export.

**Within the platform** — click **Share** to control who can open the report. Add individual users or whole roles, each at **read** (view only), **write** (edit but not delete), or **full** (everything, including delete) access. The **General access** setting controls the report's share link: **Restricted** (only the owner and the people or roles added), **Workspace** (anyone with access to the workspace), **Organisation** (anyone in your organization), or **Anyone with the link** (no account needed). A new report starts Restricted. Anyone with access can view and navigate the evidence links.

***

## Suggested structure

Generated reports follow a standard pattern. For manually built reports, the same structure works well:

```
# Executive Summary
## Key Findings
## Recommendations

# Analysis
## Data Overview
## Causal Relationships
## Simulation Results

# Appendix
## Methodology
## Data Sources
```

Lead with conclusions — executives read the top. Provide supporting detail below for those who want it.