Gate 4 panel guide
Gate 4 (hour 268) decides whether a person's capability meets the programme's standard. Completion also needs the Lean Six Sigma Green Belt that follows (Module 17, Evidence 17). The panel reviews the full capability self-assessment, the Part D practical, and evidence Evidence 14 to 16.
The panel
Three people:
- Chair: a lead test engineer or test manager at least one band above the person, and not their mentor. For Band 7 tracks, the chair is the head of test or a lead from outside the person's area.
- A developer from another team.
- The training lead.
The clinical safety officer reviews the traceability parts of Evidence 14 and Evidence 16 in writing, before the panel meets.
Each panel member needs about 3 hours per person: 1.5 hours reading beforehand, 1 hour for the session, and 30 minutes to agree and record the result.
Before the session
The training lead sends the panel, at least 5 working days before:
- the person's track guide (
materials/tracks/<track>/index.md) - the completed and calibrated instrument for Gate 4, and the Gate 3 one for comparison
- Evidence 14 (safe and lawful automation), Evidence 15 (quality engineering), and Evidence 16 (capstone)
- track module evidence: Coaching others in automation log, Automation strategy and metrics work, Acceptance test automation or Leading teams through automation adoption outputs, as the track requires
- the clinical safety officer's written review of traceability
- the product owner's acceptance of the capstone
- the individual learning plan and earlier gate review forms
- reasonable adjustments to apply.
The session (60 minutes)
| Time | Item |
|---|---|
| 5 minutes | Chair welcomes the person, explains the session and that it is developmental only. |
| 10 or 20 minutes | Capstone presentation, aimed at a non-technical audience: 10 minutes for Band 3 and Band 4, 20 minutes for Band 5 to Band 7. |
| 15 minutes | Part D practical: live run, explanation, and one small change the panel asks for (see Part D practicals). |
| 15 to 25 minutes | Questions (see below). |
| 5 minutes | The person's own reflection: what changed most, and what they will work on next. |
Questions to ask
Choose from these, tuned to the track:
- Automation choices (Learning outcome 1): Which manual tests did you keep manual, and why?
- Pyramid: Which test did you move below the UI, and what did that gain?
- Real assertions (Principle 12): Show me a test that would fail if this behaviour were wrong.
- Spec and code (Principle 12): Where does your
spec/index.mddescribe this test? - CI (Learning outcome 7): Tell me about a CI failure you triaged. Was it the product, the test, or the environment?
- Safety (Learning outcome 8, Learning outcome 9): Which hazard does this test protect? What would a clinical safety officer see as evidence?
- Data: How do you know there is no real patient data or secret in this repository?
- Flaky tests (Learning outcome 10): What would you do if this test started failing one run in ten?
- AI (Principle 15): Did you use an AI assistant? Explain this line it helped with.
- Role, not just automation (Learning outcome 13): Pick one Part B item you moved from Not yet to Meets. What do you now do differently?
- Band 6 and Band 7: How did you coach others (Coaching others in automation)? What did they learn?
- Band 6 quality assurance: How were acceptance checks agreed with clinical users (Acceptance test automation)?
- Band 7 test engineering: How do others use your framework? What did the performance test tell you about clinical demand?
- Band 7 test management: Which metric would you use to decide whether automation is working, and what would you do if it got worse?
Deciding
The panel checks every Gate 4 threshold:
- Parts A, B, and C each at least 90%, using agreed ratings
- Part D: Meets
- No skill more than one level below expected
- Test engineering at the automation target for the track
- Capstone accepted by the product owner
- Overall capability index at least 90% (Learning outcome 13)
The panel may change an agreed rating only where the session shows clear evidence against it. It records each change and its reason on the gate review form.
- All thresholds met: the person goes on to Module 17, the Lean Six Sigma Green Belt, and completes the programme when they hold the certification with an accepted Green Belt project (Evidence 17). Gate 5, about six months after Gate 4, confirms the capability holds.
- Not met: up to 22.5 extra learning hours on the items below threshold, with a written plan and extra mentor time, and one repeat of the panel. If the repeat is not met, the person, line manager, and training lead agree a next step.
What the panel does not decide
- The panel does not decide band, pay, or role.
- A panel result never starts a capability, performance, or conduct procedure.
Feedback
The chair gives the person spoken feedback on the day, with at least two strengths and the top gaps, and written feedback within 5 working days.
View this document on GitHub materials/gates/gate-4-panel-guide.md