改KaizenReports
Sign inGet started

Test plans & executions

A test plan is a curated bundle of test cases (manual, automated, or both) you want to execute together — for a release, a regression sweep, an exploratory session, or a UAT cycle.

An execution is one instance of running a plan. A plan can have many executions over time.

Creating a plan

From the project's Test Plans tab:

  1. New plan → enter a name + optional description
  2. Add cases — query the case inventory by suite, label, owner, or free text. Multi-select adds them to the plan
  3. Set ownership — who's responsible for executing each case. Defaults to the case's CODEOWNERS-derived owner
  4. Save — the plan becomes available to schedule

Plans are versioned. Editing a plan creates a new revision; existing in-progress executions stay pinned to the revision they started against.

Running a plan

From a plan, click Start execution:

  • Manual mode — Kaizen presents each case as a checklist. Reviewers mark pass / fail / skip / blocked, attach screenshots, leave notes
  • Automated mode — Kaizen sends a webhook to your CI to kick off the matching test runs. As Kensho reports stream back via kensho push, the execution auto-updates
  • Hybrid — manual cases stay manual, automated cases auto-fill from runs in the same execution window

Live execution status is shown in the project's Live Runs tab.

Auto-fill from runs

The platform's auto-fill loop watches incoming Kensho runs. If a run contains a case that's part of an open execution (matched by stable ID), the execution row auto-flips to the run's status — no manual action needed.

Disable per-execution if you want strictly manual sign-off.

Closing an execution

When all cases have a verdict (or you accept partial coverage), click Close. The execution becomes immutable history.

A summary view shows:

  • Pass / fail / skip counts
  • Coverage vs plan size
  • Average case duration
  • Defects opened during the execution

Tips

  • Smoke plan — seed a plan with @smoke labeled cases, schedule it on every release branch
  • UAT plan — manual-only plan for the QA team, run pre-release
  • Regression plan — full automated suite, scheduled nightly with auto-fill on
  • Exploratory plan — empty plan with notes only; reviewers add cases as they discover bugs