SharePoint

Every finished test run writes a standalone report.html into its run folder. This integration takes that file and puts it in a SharePoint document library, so the people who never open this tool — and the CI job that has no UI at all — find it where they already look for documents.

Configure it in Settings › SharePoint.

What it signs in as

Uploads happen as an application, not as you: that is what lets an unattended --view run upload its own report with nobody at the keyboard.

1. Register an application in Microsoft Entra ID (Azure AD).
2. Give it the Sites.ReadWrite.All application permission, and grant it admin consent.
3. Create a client secret for it, and copy the secret value.

The Directory (tenant) id and Application (client) id are on the app's overview page.

Where the report lands

• Site URL — the address of the site as the browser shows it, without the library or the folder: https://your-company.sharepoint.com/sites/QA.
• Document library — empty means the site's default library (Documents).
• Folder — inside the library. Missing folders are created on the way.
• File name — a file already there is replaced.

The folder and the file name can both be written in terms of the run: {runId}, {status}, {environment}, {trigger}, {date} and {time}. So Test reports/{environment} with {runId}-{status}.html files a failed staging run as:

Documents/Test reports/staging/2026-08-20_14-30-05-staging-failed.html

When decides which runs are uploaded: every one of them, or only the ones with a failing flow.

Where the secret lives

The client secret is the one setting that does not go into config/sharepoint.json. It is written to a .env file at the root of your context folder, as SHAREPOINT_CLIENT_SECRET — and .env is added to the context's .gitignore for you, so a context that is a git repository cannot commit it by accident. Everything else stays in the config file, which is safe to share and to review.

You can also write the variable yourself instead of typing it in the UI; the tool reads it back from that same file.

What happens when it fails

Nothing about the upload can change the outcome of a run: a report that could not be delivered leaves the run exactly as it was, and shows up as a Report upload failed badge on the run page, with the reason on hover. The report is always written into the run folder first, whether it is uploaded or not.