[feat.lab.Omega.D8+] make the full product shell walkable before capability expansion

Separating route-complete shells from capability-complete modules lets the product evolve view by view without hiding planned science behind disabled navigation or presenting it as shipped.

Author: ivanicu

.github/workflows/node.yml

@@ -10,14 +10,14 @@ jobs:
   build:
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v2
-    - uses: actions/setup-node@v2
+    - uses: actions/checkout@v4
+    - uses: actions/setup-node@v4
       with:
         node-version: 22
     - run: npm ci
     - run: sudo apt-get install xvfb
-    - name: Lint
-      run: npm run lint
+    - name: Lint debt ratchet + clean product shell
+      run: node scripts/check_lint_ratchet.mjs
     - name: Test
       run: npm install --no-save "gl@^6.0.2" && xvfb-run --auto-servernum npm run jest
     - name: Build

ARCHITECTURE.md

@@ -96,8 +96,8 @@ current schema history.
 ## Client application
 
 `src/app/shell/registries.ts` is the one registry for eight Workspaces, thirty Views and
-composable modules. Gated Views exist for information architecture but only implemented
-Views are surfaced. `AppShell` owns routing and deep-link restoration;
+composable modules. Every View has a stable route and product shell; `implemented` remains
+the separate truth for connected scientific capability. `AppShell` owns routing and deep-link restoration;
 `ScientificContextStore` owns Program, focus, selection and the application's only
 staleness generation; `SceneService` owns one mol* instance that survives navigation.
 
@@ -147,7 +147,8 @@ rebuilds. The model can therefore calibrate observed command latency and outcome
 addition to detecting drift, ranking static hotspots and estimating dependency radius.
 It still cannot predict unseen inputs or autonomously change the architecture; that is
 the L4 boundary. The platform substrate is complete against its approved DoD; the
-product is explicitly partial (currently 3/8 Workspaces and 7/30 Views implemented).
+product capability is explicitly partial (currently 3/8 Workspaces and 7/30 Views connected)
+even though the navigable product shell is complete at 8/8 and 30/30.
 
 `scripts/digital_twin_scope.json` is the ownership boundary. The watcher recursively
 discovers tracked and untracked files under all first-party roots and automatically

STATUS.md

@@ -13,7 +13,7 @@ Last re-derived: **2026-08-12**, from the running local system and repository ga
 | Durable execution | PostgreSQL JobStore, bounded ThreadExecutor, plus Inline/Process/Remote executor contracts |
 | Results | content-addressed artifacts; method-current field cache and generic deterministic result cache |
 | Application context | one ScientificContextStore generation; LigandStore and async facets delegate to it |
-| AppShell | exactly 8 Workspaces and 30 Views; only implemented Design, Structures and Runs views surfaced; ModuleHost composes runtime modules from the registry |
+| AppShell | exactly 8 Workspaces and 30 Views; all have navigable, human-readable shells while connected capability remains explicit; ModuleHost composes runtime modules from the registry |
 | 3D scene | one SceneService-owned mol* instance survives route navigation |
 | Durable domain | Program, Campaign, Hypothesis, Evidence, Decision, Mission, Run, Run↔Job and controlled relations |
 | Attention | derived from actionable Job outcomes and approval-waiting Runs; exposed through the same `attention.list` command boundary |
@@ -46,7 +46,8 @@ Last re-derived: **2026-08-12**, from the running local system and repository ga
 
 - All 18 architecture gate checks green, including build, contracts, migrations, parity,
   layering, security and the digital twin.
-- TypeScript typecheck and focused AppShell/ScientificContext/ModuleHost specs: green.
+- TypeScript typecheck and focused AppShell/ScientificContext/ModuleHost specs: green; all
+  30 View routes are restorable and skeleton Views unmount operational modules cleanly.
 - Production `dirac` bundle and real deep-link browser flow: green; `/p/KRAS-G12D/structures/complex`
   restored Program + Complex + Molecule and a live Mol* scene.
 - Real browser navigation `Structures → Design → Structures → Runs` preserved the exact
@@ -70,12 +71,13 @@ Last re-derived: **2026-08-12**, from the running local system and repository ga
 
 The default local/LAN profile intentionally remains unauthenticated. Public operation must
 explicitly activate the documented remote profile behind HTTPS with real operator-issued
-credentials. Product expansion Workspaces remain gated until their scientist-facing views
-are implemented; the platform architecture they will use is now present.
+credentials. All 8 Workspaces and 30 Views now have stable routes and honest product shells;
+planned modules show their real backing state instead of placeholder scientific results.
 
 ## Scope verdict
 
-The approved **platform substrate phase is complete**. The whole Dirac product is not:
-3 of 8 Workspaces and 7 of 30 Views currently expose implemented vertical slices. The
-remaining registry entries are product intent protected by capability gates, not shipped
-features. The Architecture Optimization Twin reports this distinction directly.
+The approved **platform substrate and product-shell phases are complete**. The whole Dirac
+product is not: 8 of 8 Workspaces and 30 of 30 Views are navigable, while 3 of 8 Workspaces
+and 7 of 30 Views currently expose connected scientific vertical slices. The remaining
+Views are implementation contracts with explicit foundation/planned states, not shipped
+scientific capability. The Architecture Optimization Twin reports both dimensions.