Skip to content

Reports & failed tests

Runs are listed by run number from .robotstudio/reports/Run-* (legacy root Reports/ paths remain readable when stored). Opening Reports selects the latest run automatically when any exist, so the details page is ready without waiting on a click. The active run in the sidebar is highlighted so you can see which details are open.

Each run entry shows stats and links to the artifacts Robot Framework produced. The list badge is PASS / FAIL from test results. A Robot exit with no tests (nothing matching a tag, empty suite, exit 252) is NO TESTS — not FAIL — and does not count toward pass rate or Recent Failures. A framework crash is ERROR. Failure counts stay emphasized; there is no pass/fail status badge on the details header. Run details also show the configuration used for that run (Default when none was selected).

  • report.html
  • log.html
  • output.xml

Generate ReportLens in the Reports header (next to Refresh) builds a modern interactive HTML report from the selected run’s output.xml (on demand — not after every run) and opens it in your default browser. The file is written as reportlens.html beside the other artifacts and reused until output.xml changes. The button stays disabled until a run with output.xml is selected.

Delete Run on the details panel asks for confirmation, then removes:

  • The .robotstudio/reports/Run-* folder (report.html, log.html, output.xml, and any reportlens.html)
  • That run’s history row from Studio’s database (~/.robot-studio/robot-studio.db)

Insights uses the same history, so the run drops out of pass rate, last run, streaks, and per-file stats on the next Insights load. If Insights is already open, use Refresh.

Under Settings → Execution → Report Retention Days, set how long to keep run folders. 0 keeps them forever. A positive value deletes runs older than that many days when you open a workspace or save Settings.

Important: Auto-deleted runs are removed from Insights analytics as well (same effect as Delete Run). Use retention only when you are comfortable losing that history to reclaim disk space.

Failed runs surface a Failed Tests section on the run details page (same data as Live Execution):

  • Failure message from Robot
  • Source location
  • Jump to Source — open the failing test at the right place
  • Re-run Test — execute just that failure again

After a live run finishes, the same list also appears on the Execution page.

When a run finishes, stay in context and open the report from the toast when you are ready. Clickable run status in the chrome returns you to Tests if you want the suite tree again.

From Insights, Last run, fail streak, headline Failed, and file triage Open failed tests / View report land on this same Reports view.