Skip to content
Robot StudioRobot StudioRobot Studio

Insights

Insights is a triage starting point for run health — not a metrics dashboard. Open it from the activity bar (Insights) or the command palette.

Use it when you want to know where the project is hurting and jump straight into Failed Tests, source, Reports, or a rerun.

Run health is computed from the same run history as Reports. Deleting a run removes it from that history (database row and Run-* folder), so Insights no longer counts it. The same applies when Report Retention Days auto-deletes old runs. If Insights is already open, use Refresh.

Run health

Section titled “Run health”

Once you have executed suites, the left panel (top on a narrow window) is the primary surface. Numbers are about runs (each time you pressed Run), not individual test cases — so a suite with 4 passes and 2 fails still counts as one Failed run.

Project / Tag / multi-select runs still count once in the headline strip. In the Files table they are broken out per .robot file from that run’s saved file outcomes (file_outcomes.json, built from output.xml when missing), so a full-project pass shows Pass = 1 on each file that ran (not zero).

Headline strip

Section titled “Headline strip”
Label What you are seeing
Pass rate Share of PASS vs FAIL runs. Empty suites (NO TESTS), framework crashes (ERROR), cancelled, and aborted runs are left out of this rate.
Runs How many runs Insights is counting from history (same store as Reports).
Failed How many of those runs failed. Tap to open the last failed run in Reports.
Avg duration Average wall time of runs that recorded a duration.

Health cards and charts

Section titled “Health cards and charts”
Signal What you are seeing
Pass streak / Fail streak How many newest runs in a row share the same pass or fail outcome. Tap to open that latest run in Reports.
Flaky files Suites that have both passed and failed across history — unstable enough to triage.
Interrupted Runs you stopped, or that never fully started.
Pass rate trend How the recent pass rate has moved over the last few runs.
Duration trend How long recent runs took (peak and average called out under the chart).
Last run The newest run’s outcome and timing. Tap to open it in Reports.
Failure mix by suite Which files account for the most failed runs. Tap a bar to open file triage.

A short Last N note may appear when the newest handful of runs looks much better or worse than the overall pass rate — useful when history is long but recent health changed.

File triage

Section titled “File triage”

Click a failing file in Failure mix or the Files table (for example login.robot — 9 failed / 21 runs). The triage dialog shows failure counts and the last failed test name when it is known, then:

Action Destination
Open failed tests Reports, on the last failed run for that file
Open source The suite in the editor
View report Reports for the last failed run, or the last run if none failed
Rerun file Execution, running that .robot file

Files with no failures still open in the editor.

Composition

Section titled “Composition”

Composition is secondary — indexed shape of the project, not execution results.

After the index is ready:

Metric Meaning
Keywords / test cases / variables / … Definitions found by the indexer
Suites Distinct .robot files
Files Distinct indexed source files (.robot, .resource, .py, …)

Use Rebuild Index if counts look stale after adding files.

Variables counts user-declared names from:

  • *** Variables ***
  • [Arguments]
  • VAR
  • ${x}= … assignments
  • FOR loop variables
  • YAML variable files (when indexed)

It does not count BuiltIn automatic variables (${TEST_NAME}, …) or environment %{ENV} references.

Keywords counts user keyword definitions under *** Keywords ***, plus public methods from indexed project .py libraries — not BuiltIn keywords and not call sites.

Also in Composition:

Section Meaning
Composition mix Stacked bar of keywords / tests / variables
File types Counts of .robot, .resource, .py, .yaml among indexed files
Focus Density chips, index state, densest files, and test-heavy files

Focus details:

  • Density chips — keywords per test, symbols per file, resource and library counts
  • Index state
  • Densest files — files with the most indexed definitions (click to open)
  • Test-heavy files — files with the most test-case definitions (click to open)

Kind rows with a count of zero stay visible but muted so small projects still show the full shape of the index.

Files table

Section titled “Files table”

The Files table lists executed .robot suites only — helpers like .py / .resource and suites that have never run stay out. Indexed KW / TC / Var counts still appear next to run health. Rows load as you scroll.

Column Meaning
File Suite path
KW / TC / Var Indexed keyword, test-case, and variable definitions in that file
Runs Times that file was part of a run
Pass / Fail How many of those runs passed or failed overall
Stop Runs that did not finish — you pressed Stop, or Robot never started (aborted launches are not kept in Reports)
Last Outcome of the newest run that touched this file

Failing rows open triage; clean rows open the file in the editor.

First open on a large project

Section titled “First open on a large project”

The first Insights open after launching a very large suite can take longer while composition and per-file run stats are computed. The spinner advances short status lines every 30 seconds while you wait. If it fails, use Refresh — a second open is usually fast once caches are warm.

Section titled “Related”