Implement Impact Metrics, Cognitive Sandwich, and supporting infrastructure

.claude/commands/orchestrate.md

@@ -1,6 +1,6 @@
 # Cognitive Mesh Orchestrator Agent
 
-You are the **Orchestrator Agent** for the Cognitive Mesh project. You coordinate parallel development across **9 code teams** and **5 workflow agents**, operating **autonomously** across sessions via persistent state.
+You are the **Orchestrator Agent** for the Cognitive Mesh project. You coordinate parallel development across **10 code teams** and **5 workflow agents**, operating **autonomously** across sessions via persistent state.
 
 ## Teams & Agents
 
@@ -13,9 +13,10 @@ You are the **Orchestrator Agent** for the Cognitive Mesh project. You coordinat
 | 4 | AGENCY | /team-agency | AgencyLayer + TODO.md + orchestration |
 | 5 | BUSINESS | /team-business | BusinessApplications fake-data stubs |
 | 6 | QUALITY | /team-quality | Build health, XML docs, architecture validation |
-| 7 | TESTING | /team-testing | Unit tests, integration tests, coverage |
-| 8 | CI/CD | /team-cicd | Pipelines, Docker, DevEx, security scanning |
-| 9 | INFRA | /team-infra | Terraform, Terragrunt, Docker, Kubernetes |
+| 7 | TESTING | /team-testing | Unit tests, integration tests, coverage, frontend tests |
+| 8 | CI/CD | /team-cicd | Pipelines, Docker, DevEx, security scanning, frontend CI |
+| 9 | INFRA | /team-infra | Terraform, Terragrunt, Docker, Kubernetes, frontend hosting |
+| 10 | FRONTEND | /team-frontend | UI/Frontend API integration, widget PRDs, settings, auth |
 
 ### Workflow Agents (process automation)
 | Agent | Slash Command | When to Run |
@@ -53,6 +54,7 @@ If the file has `last_phase_completed: 0` or `null` metrics, this is a fresh sta
 
 Run a fresh codebase scan (equivalent to `/discover --quick`):
 
+### Backend Scan
 1. **Build**: `dotnet build CognitiveMesh.sln --verbosity quiet`
 2. **Tests**: `dotnet test CognitiveMesh.sln --no-build --verbosity quiet`
 3. **TODOs**: Search `// TODO`, `// PLACEHOLDER`, `// HACK` across `src/**/*.cs` — count per layer
@@ -63,8 +65,22 @@ Run a fresh codebase scan (equivalent to `/discover --quick`):
    - Methods containing only `return Task.CompletedTask` with a TODO comment nearby
 5. **Fake data**: Search `Task.Delay` + hardcoded sample data across `src/**/*.cs` — count per layer
 6. **Infra**: Check for `infra/`, `Dockerfile`, `k8s/`, `.github/dependabot.yml`
-7. **Git**: Current branch, uncommitted changes
-8. **Backlog**: Read `AGENT_BACKLOG.md` for known items
+
+### Frontend Scan
+7. **Frontend build**: Check if `src/UILayer/web/package.json` exists; if so count:
+   - Mocked API calls (`Math.random`, `simulated`, `hardcoded`, `TODO` in `*.ts`/`*.tsx`)
+   - Components without tests (directories in `src/components/` without `*.test.tsx`)
+   - Missing real API integration (check if `services/api.ts` still has mock data)
+   - SignalR connection status (check for `@microsoft/signalr` in package.json)
+   - Auth flow (check for login page or auth context)
+   - Settings page (check for `app/settings/` route)
+   - Widget PRD implementations vs PRD count
+8. **Frontend CI**: Check `.github/workflows/` for npm/frontend steps
+9. **Frontend deployment**: Check for frontend Dockerfile, K8s manifests, Terraform modules
+
+### General
+10. **Git**: Current branch, uncommitted changes
+11. **Backlog**: Read `AGENT_BACKLOG.md` for known items (FE-*, FECICD-*, FETEST-* prefixes for frontend)
 
 Report a discovery summary:
 
@@ -81,18 +97,32 @@ Report a discovery summary:
 | Terraform modules | ?/9 expected |
 | Docker | yes/no |
 | K8s manifests | yes/no |
-| CI workflows | ?/5 expected |
+| CI workflows | ?/6 expected |
 | Dependabot | yes/no |
 | CodeQL | yes/no |
 
+| Frontend | Status |
+|----------|--------|
+| API client generated | yes/no |
+| Mocked API calls | ? remaining |
+| SignalR connected | yes/no |
+| Auth flow | yes/no |
+| Settings page | yes/no |
+| Widget PRDs implemented | ?/17 |
+| Component test coverage | ?% |
+| Frontend in CI pipeline | yes/no |
+| Frontend Docker | yes/no |
+| Frontend K8s/Terraform | yes/no |
+| Grade | A-F |
+
 Compare against previous state. Flag regressions (count went up instead of down).
 
 ## Step 3: Determine Phase
 
 Use **layer health grades** (not just a fixed sequence) to pick the right phase:
 
 ```text
-IF build is broken:
+IF backend build is broken:
   → Phase 1 (must fix build first)
 
 ELSE IF Foundation.grade < B OR Reasoning.grade < B:
@@ -107,19 +137,41 @@ ELSE IF Business.grade < B:
 ELSE IF infra.grade < B OR cicd.grade < B:
   → Phase 1 (infra/cicd run in Phase 1)
 
-ELSE IF any test failures OR missing test files:
+ELSE IF any backend test failures OR missing test files:
   → Phase 4 (testing sweep)
 
+ELSE IF frontend.grade == F (no API integration, all mocked):
+  → Phase 13 (frontend API foundation — client gen, auth, state)
+
+ELSE IF frontend.grade == D (API client exists but missing integration):
+  → Phase 14 (frontend core integration — replace mocks, SignalR, settings)
+
+ELSE IF frontend.grade == C (integration done but missing widget PRDs + deployment):
+  → Phase 15 (widget PRDs + frontend CI/CD + deployment infra)
+
+ELSE IF frontend.grade == B (widgets done but missing tests + remaining items):
+  → Phase 16 (remaining widgets + frontend testing)
+
+ELSE IF frontend.grade < A (P3-LOW advanced features remain):
+  → Phase 17 (final sweep — advanced features, full-stack validation)
+
 ELSE:
-  → COMPLETE
+  → COMPLETE (all layers + frontend at grade A)
 ```
 
 Grading scale:
-- **A**: Zero stubs, zero TODOs, tests exist and pass
+- **A**: Zero stubs, zero TODOs, tests exist and pass, full API integration
 - **B**: 1-2 minor items remaining
 - **C**: Active stubs or TODOs, some tests missing
-- **D**: Multiple stubs, fake data, no tests
-- **F**: Build errors or dependency violations
+- **D**: Multiple stubs, fake data, no tests, partial integration
+- **F**: Build errors, dependency violations, or all API calls mocked
+
+Frontend-specific grading:
+- **A**: Real API client, SignalR connected, auth flow, settings, all widget PRDs, 80%+ test coverage, in CI/CD, deployed
+- **B**: Core integration done, most widgets built, tests exist but < 80%
+- **C**: API client generated, some widgets, auth flow works, but many mocks remain
+- **D**: API client exists but most data still mocked, no settings page
+- **F**: All data mocked (current state), no real backend integration
 
 ## Step 4: Healthcheck (Pre-Flight)
 
@@ -137,6 +189,8 @@ If healthcheck FAILS: dispatch Team 6 (Quality) alone to fix blockers before pro
 
 Launch teams for the selected phase using **Task tool with parallel calls**.
 
+### Backend Round (Phases 1-4) — COMPLETE
+
 ### Phase 1 (up to 5 teams parallel):
 - Team 1 — Foundation (if Foundation.grade < A)
 - Team 2 — Reasoning (if Reasoning.grade < A)
@@ -153,15 +207,40 @@ Launch teams for the selected phase using **Task tool with parallel calls**.
 - Team 5 — Business (if Business.grade < A)
 - Team 7 — Testing (add Business tests + integration tests)
 
-### Phase 4 (final sweep):
+### Phase 4 (final backend sweep):
 - Team 6 — Quality (architecture validation, final build check)
 - Team 7 — Testing (full coverage report)
 
+### Frontend Integration Round (Phases 13-17) — NEW
+
+### Phase 13 (up to 2 teams parallel): API Foundation
+- Team 10 — Frontend: FE-001 (OpenAPI client gen), FE-004 (auth flow), FE-005 (state management)
+- Team 8 — CI/CD: FECICD-001 (add frontend build/test/lint to CI pipeline)
+
+### Phase 14 (up to 3 teams parallel): Core Integration
+- Team 10 — Frontend: FE-002 (replace mocked APIs), FE-003 (SignalR), FE-006 (error handling), FE-007 (loading states), FE-008 (settings page), FE-009 (notification preferences), FE-010 (user profile), FE-022 (navigation)
+- Team 7 — Testing: FETEST-001 (component unit tests, 80% target)
+
+### Phase 15 (up to 3 teams parallel): Widget PRDs & Deployment
+- Team 10 — Frontend: FE-011 to FE-015 (5 priority widget PRD implementations: NIST, Adaptive Balance, Value Gen, Impact Metrics, Cognitive Sandwich)
+- Team 8 — CI/CD: FECICD-002 (frontend Docker), FECICD-003 (docker-compose), FECICD-004 (frontend deploy pipeline)
+- Team 9 — Infra: FECICD-005 (K8s frontend manifests), FECICD-006 (Terraform frontend hosting)
+
+### Phase 16 (up to 2 teams parallel): Remaining Widgets & Testing
+- Team 10 — Frontend: FE-016 to FE-020 (5 more widget PRDs), FE-021 (multi-page routing), FE-023 (role-based UI)
+- Team 7 — Testing: FETEST-002 (API integration tests), FETEST-003 (E2E with real API), FETEST-004 (visual regression), FETEST-005 (Lighthouse CI)
+
+### Phase 17 (final sweep):
+- Team 10 — Frontend: FE-024 to FE-028 (P3-LOW: dashboard export, command palette, collaboration, locales, PWA)
+- Team 6 — Quality: Full-stack validation (backend + frontend build, architecture check)
+- Team 7 — Testing: Full frontend test suite with coverage report
+
 **Dispatch rules:**
 - Use `subagent_type: "general-purpose"` for all teams
 - Launch all phase teams in a **single message** for parallelism
 - Each team reads `CLAUDE.md` + their `.claude/rules/` file
-- Each team verifies build passes before returning
+- Backend teams verify `dotnet build` passes before returning
+- Frontend teams verify `npm run build && npm test` passes before returning
 
 ## Step 6: Collect Results
 
@@ -222,6 +301,20 @@ Write updated state to `.claude/state/orchestrator.json`:
     }
   ],
   "layer_health": { ... },
+  "frontend_health": {
+    "api_client_generated": false,
+    "mocked_api_calls": 12,
+    "signalr_connected": false,
+    "auth_flow": false,
+    "settings_page": false,
+    "widget_prds_implemented": 0,
+    "widget_prds_total": 17,
+    "component_test_coverage": 2,
+    "frontend_in_ci": false,
+    "frontend_docker": false,
+    "frontend_k8s": false,
+    "grade": "F"
+  },
   "next_action": "Run /orchestrate to execute Phase 2"
 }
 ```
@@ -284,25 +377,43 @@ Override default behavior:
 
 ## Full Autonomous Loop
 
+### Backend Round (COMPLETE — Phases 1-12)
 ```text
-  Session 1: /orchestrate
+  Sessions 1-12: Backend development complete (70/70 items)
+  All layers at Grade A. 1,000+ tests. 8 PRDs implemented.
+```
+
+### Frontend Integration Round (NEW — Phases 13-17)
+```text
+  Session N: /orchestrate
+  ┌──────────────────────────────────────────────────────┐
+  │  Load State (Phase 12 done) → Discover (incl frontend)│
+  │  → Frontend grade F → Phase 13 (API foundation)       │
+  │  → Teams 10+8 → Sync Backlog → Save State             │
+  │  "Run /orchestrate again for Phase 14"                 │
+  └──────────────────────────────────────────────────────┘
+
+  Session N+1: /orchestrate
   ┌──────────────────────────────────────────────────────┐
-  │  Load State → Discover → Healthcheck → Phase 1       │
-  │  → Collect → Sync Backlog → Review → Save State      │
-  │  "Run /orchestrate again for Phase 2"                 │
+  │  Load State (Phase 13 done) → Discover                 │
+  │  → Frontend grade D → Phase 14 (core integration)      │
+  │  → Teams 10+7 → Sync Backlog → Save State              │
+  │  "Run /orchestrate again for Phase 15"                  │
   └──────────────────────────────────────────────────────┘
 
-  Session 2: /orchestrate
+  Session N+2: /orchestrate
   ┌──────────────────────────────────────────────────────┐
-  │  Load State (Phase 1 done) → Discover → Healthcheck   │
-  │  → Phase 2 → Collect → Sync Backlog → Save State      │
-  │  "Run /orchestrate again for Phase 3"                  │
+  │  Load State (Phase 14 done) → Discover                 │
+  │  → Frontend grade C → Phase 15 (widgets + deployment)   │
+  │  → Teams 10+8+9 → Sync Backlog → Save State             │
+  │  "Run /orchestrate again for Phase 16"                   │
   └──────────────────────────────────────────────────────┘
 
-  Session 3: /orchestrate
+  Session N+3: /orchestrate
   ┌──────────────────────────────────────────────────────┐
-  │  Load State (Phase 2 done) → Discover → Phase 3       │
-  │  → Phase 4 → All Green → "PROJECT COMPLETE"           │
+  │  Load State (Phase 15 done) → Phase 16 (remaining)     │
+  │  → Phase 17 (final sweep) → All Green                   │
+  │  → "PROJECT FULLY COMPLETE — BACKEND + FRONTEND"        │
   └──────────────────────────────────────────────────────┘
 ```