Skill 71 · Setting Up Regression Runs
Subchapter 71.1
references/regression-configurations.mdMarkdown2 KBView on GitHub
These platforms ship several same-named runs distinguished only by Testmo Configuration. The API can only create step 1 of each; the step 2 UI pass is explained in the skill’s , and each spec’s gives its exact pass.
bitwarden-testmo-toolscreating-test-runsreferences/multi-configuration-runs.md_commentMobile. Same filters, differing only by which automation type is
excluded. config_id 1 = Android, 3 = iOS.Desktop and Extension use a broad + narrow split: one variant covers the full folder and the rest are driven by an “essential” tag.
Desktop. Broad = macOS (21); narrow = Windows (22), Linux
(20), driven by the desktop-essential tag (id 15999).Extension. Broad = MacOS, Chrome (7); narrow =
Windows, Edge (15), MacOS, Safari (10), Linux, Firefox (5), driven by the
extension-essential tag (id 17190).Config and tag ids above verified 2026-08-28 via GET /projects/1/configs and GET /projects/1/tags.
Every variant on both platforms, broad and narrow alike, takes a step 2. The two variant types differ only in how cases are selected:
| Broad | Narrow | |
|---|---|---|
| Scope | whole top-level folder + subfolders | the <platform>-essential tag (server-side) |
| Test Type | Smoke + Regression | any — the tag defines the scope |
| Case state | Active | Active |
| Automation | manual only (exclude 10, 23, 24) | manual only (exclude 10, 23, 24) |
Narrow variants deliberately drop the test-type filter, since a tagged case counts regardless of type, but keep the state and automation-type filters so retired and already-automated cases stay out.