Skip to main content

Quality Workspace

The dashboard Quality workspace is the portfolio-level entry point for Fairway's cited Quality Record. It answers a different question from the other dashboard views:

ViewPrimary question
WallWhat is active or blocked now?
BoardWhich tasks are moving through the configured workflow?
QualityWhat lifecycle evidence exists for each bounded work item?
ReportsWhat changed during a delivery window?
ControlsDo specific recorded controls discriminate observable outcomes?
DiagnosticsIs the coordination/runtime machinery healthy?

Portfolio Projection​

/quality shows one row per visible task and one column per Quality Record stage:

  1. intent;
  2. material decisions;
  3. production context;
  4. collected evidence;
  5. automatic verification;
  6. human judgment;
  7. promotion;
  8. operational outcomes; and
  9. lessons and controlled improvement.

The workspace preserves the canonical states present, missing, unavailable, conflicting, and externally_owned. It does not collapse them into a score. Every stage cell links to the task-level Quality Record, where the underlying fact and source references remain inspectable.

The default view is bounded and paginated. Search, task status, role, profile, and risk filters are applied before the visible page is projected. Global session and checkpoint context is loaded once per page projection, and the result participates in the dashboard's short-lived read-model cache.

Interpretation​

  • missing means an expected Fairway record is absent.
  • unavailable means Fairway lacks an authoritative fact; it is not evidence that no event or outcome occurred.
  • conflicting retains contradictory evidence for accountable interpretation.
  • externally_owned identifies authority retained by Git, CI/CD, deployment systems, reviewers, operators, or another named system.

The workspace is read-only. It cannot approve, waive, merge, deploy, release, accept risk, or mutate workflow. Control-effectiveness analysis remains on the Controls page so lifecycle completeness and statistical discrimination are not presented as the same claim.