Agree on the rules.
Keep the evidence visible.
We have a working review process. The next step is deciding how the team wants it to handle incomplete measurements, system priorities, and new evidence.
Where the project stands
- All 155 original records were rerun with the repository connected. Source Pass/Fail and condition grades stayed unchanged; 74 remain Unscored.
- 44 urgency categories changed. High priority went from 48 to 62; urgent remained two. Nine of the previous top 20 remain in the top 20.
- The order changed. This is not validated improvement and does not isolate the effect of repository content. Specialists still need to review whether the changes make sense.
When a reading cannot be taken
First establish whether the measurement applies. If it does, use a valid reading when available. If it cannot be obtained, record why.
- Unavailable and not applicable are different states.
- Use an alternative only when its rule and required evidence are approved.
- Continue independent checks. Do not invent a replacement or calculate dependent results from a missing reading.
- If no valid route remains, leave the condition Unscored and keep the record in the review queue.
Decide: Which readings can be unavailable, what evidence supports an alternative, and who resolves an unknown context?
What should system health mean?
The current tool scores individual records. A system-wide health score is still to be defined.
- Start with the decision: maintenance, retesting, or specialist review.
- Show the worst findings, testing coverage, and missing evidence beside any summary.
- An average can hide a serious problem. Define which findings must override it.
Decide: What makes a whole system a priority even when many individual readings look good?
Add evidence without mixing it up
A new survey reading and a new type of measurement need different checks.
- Keep the date, units, reference basis, method, and matching asset with the reading.
- Agree which readings can be paired; different survey dates can change their meaning.
- Map and validate new fields before scoring them. Preserve earlier results so changes can be explained.
Decide: What is the next dataset, and who confirms its measurement details before it affects the ranking?
Keep the project rules in one place
The linked wiki records the methodology, decisions, open questions, and lessons. Its editable files give the team a common reference.
- Jev receives five structured references: scoring, urgency, qualifiers, practices, and review rules.
- Each run records the guidance used. The original result remains available.
- Measurement applicability and alternate routes are proposed additions; the draft policies are not active scoring rules.
Decide: What belongs in the shared wiki, what stays private, and who maintains the rules?
Leave with a name and a next step
Assign an owner to the measurement rules, the system-health definition, the next dataset, and the wiki. Select a small set of rankings for specialist review and agree when to review the next version.
Keep separate: original Pass/Fail, calculated condition, Jev advisory urgency, and human approval. Require explanations and a chance to intervene before decisions become work plans.