Prioritized, actionable work items. Backend is 100% complete (70/70 items). Frontend Phase 13 ✓ (4/4), Phase 14 ✓ (6/6), Phase 14b ✓ (4/4). Remaining 25 frontend items + 6 DevOps evaluation tickets across Phases 15–18.
All backend items complete — click to expand
- P0-CRITICAL (3/3): Build fixes, architecture violations — all resolved
- P1-HIGH Stubs (16/16): All core stub implementations replaced with real logic
- AGN-001–006: DecisionExecutor, MultiAgentOrchestration, InMemoryAgentKnowledge, CheckpointManager, DurableWorkflow, tests
- META-001–006: SelfEvaluator, PerformanceMonitor, ACPHandler, SessionManager, LearningManager, ContinuousLearning
- FND-001–003: DocumentIngestion, EnhancedRAG, SecretsManagement
- RSN-001–004: SystemsReasoner, DomainSpecificReasoner, ValueGeneration, AnalyticalReasoner
- P1-HIGH CI/CD (8/8): CodeQL, Dependabot, Dockerfile, docker-compose, Makefile, templates, deploy pipeline, coverage
- P1-HIGH IaC (11/11): 9 Terraform modules, Terragrunt, Kubernetes manifests
- P2-MEDIUM Stubs (5/5): CustomerIntelligence, DecisionSupport, ResearchAnalyst, ConvenerController, KnowledgeManager
- P2-MEDIUM Tests (9/9): 334 new backend tests across all layers
- P2-MEDIUM PRDs (8/8): NIST, AdaptiveBalance, CognitiveSandwich, CognitiveSovereignty, TemporalDecision, MemoryStrategy, ValueGeneration, ImpactMetrics
- P3-LOW (10/10): Telemetry, performance monitoring, real-time SignalR, notifications, E2E, i18n, accessibility, code splitting, service worker, D3 visualizations
Phase 13 Foundation items complete — click to expand
- FE-001: TypeScript API client from OpenAPI — openapi-typescript + openapi-fetch, type-safe generated clients for Services and Agentic APIs
- FE-004: Auth flow — login page, JWT context with access/refresh tokens, protected route wrapper, middleware auth guard, cookie sync for SSR
- FE-006: Error handling — ErrorBoundary, Toast system with auto-dismiss, API interceptors (401→logout+redirect, 403→toast, 429→rate limit toast, 5xx→error toast)
- FECICD-001: Frontend CI job — npm ci, TypeScript check, Jest tests, coverage upload in build.yml
- Blocker resolved: AdaptiveBalance + NIST controller route decorators added
- P0-CRITICAL: Blocks all other frontend work. Do first.
- P1-HIGH: Core frontend functionality. Do in Phase 14-15.
- P2-MEDIUM: Feature completeness. Do in Phase 16.
- P3-LOW: Enhancement / polish. Do in Phase 17.
- DEVOPS: Tool evaluation and selection. Phase 18.
All items completed in Phase 13. See "Completed (Frontend Phase 13)" above.
FE-001: Generate TypeScript API Client from OpenAPI✓FE-002: Replace Mocked API Service→ Moved to Phase 14FE-003: Add SignalR Client→ Moved to Phase 14FE-004: Add Authentication Flow✓
- File:
src/UILayer/web/src/services/api.ts - Action: Replace hardcoded
DashboardAPIsingleton with real HTTP calls to all 13 backend controllers. Wire generated client from FE-001. Remove allMath.random()and simulated delays. - Depends on: FE-001 ✓
- Team: 10 (Frontend)
- File:
src/UILayer/web/src/lib/realtime/(new) - Action: Install
@microsoft/signalr, connect toCognitiveMeshHub. Replace 5-second polling interval with real-time subscriptions (JoinDashboardGroup, SubscribeToAgent). Add reconnection logic with exponential backoff. Connection state indicator in UI. - Team: 10 (Frontend)
- File:
src/UILayer/web/src/stores/(new) - Action: Add Zustand stores:
useAuthStore,useAgentStore,useDashboardStore,useNotificationStore,usePreferencesStore. Replace scattered useState with centralized stores. Add persistence middleware for preferences. - Depends on: FE-004 ✓
- Team: 10 (Frontend)
- Files:
src/UILayer/web/src/components/skeletons/ - Action: Skeleton components for dashboard panels, agent lists, metrics cards. Suspense boundaries per route. Optimistic updates for mutations (agent status changes, settings saves).
- Team: 10 (Frontend)
- Files:
src/UILayer/web/src/app/settings/ - Action: Settings page with sections: Theme (light/dark/system), Language (en-US/fr-FR/de-DE), Accessibility (reduced motion, high contrast, font size), Data & Privacy (consent toggles). Persist to backend user preferences API + localStorage fallback.
- Team: 10 (Frontend)
- Files:
src/UILayer/web/src/app/settings/notifications/ - Action: Notification preferences panel: channel toggles (email, push, SMS, in-app), category filters (approvals, security, system), quiet hours with timezone. Wire to backend Notification Preferences API.
- Team: 10 (Frontend)
- PRD:
docs/prds/01-foundational/nist-ai-rmf-maturity/mesh-widget.md - Backend:
NISTComplianceController— score, checklist, evidence, gap analysis, roadmap - Action: Maturity score gauge (D3 radial), pillar breakdown cards, evidence upload form, gap analysis table with severity indicators, roadmap timeline (Gantt-style). Wire to 7 REST endpoints.
- Team: 10 (Frontend)
- PRD:
docs/prds/02-adaptive-balance/mesh-widget.md - Backend:
AdaptiveBalanceController— spectrum, history, override, learning, reflexion, recommendations - Action: Interactive spectrum sliders (5 dimensions), real-time position updates via SignalR, override controls with approval flow modal, audit trail table, recommendation cards with accept/dismiss.
- Depends on: FE-003
- Team: 10 (Frontend)
- PRD:
docs/prds/04-value-impact/value-generation/mesh-widget.md - Backend:
ValueGenerationController— value-diagnostic, org-blindness, employability - Action: Upgrade existing TwoHundredDollarTestWidget + ValueDiagnosticDashboard. Add consent flow, scoring visualization (radar chart), strengths/opportunities display, org blindness heatmap. Wire to real API.
- Team: 10 (Frontend)
- PRD:
docs/prds/04-value-impact/impact-driven-ai/mesh-widget.md - Backend:
ImpactMetricsController— safety, alignment, adoption, assessment - Action: Psychological safety gauge (6 dimensions), mission alignment radar, adoption telemetry timeline, resistance indicator cards. 8 REST endpoints.
- Team: 10 (Frontend)
- PRD:
docs/prds/01-foundational-infrastructure/mesh-orchestration-hitl.md - Backend:
CognitiveSandwichController— create, get, advance, step-back, audit, debt - Action: Phase stepper UI (Human→AI→Human flow), HITL approval modal with diff viewer, cognitive debt tracker (burndown chart), audit log viewer with filtering. Real-time via SignalR.
- Depends on: FE-003
- Team: 10 (Frontend)
- Files:
src/UILayer/web/src/app/profile/ - Action: Account info display, role badges, GDPR consent management, data export request button, session history table. Wire to ComplianceController for consent records.
- PRD:
docs/prds/03-agentic-cognitive-systems/context-engineering-widget.md - Action: AI context frame management UI — context window visualizer, token budget, frame composition editor.
- PRD:
docs/prds/07-agentic-systems/agentic-ai-system/mesh-widget.md - Backend:
AgentController— registry CRUD, orchestrate, authority - Action: Upgrade existing AgentControlCenter. Add agent lifecycle management, authority scope configurator, orchestration trigger with real-time results streaming.
- PRD:
docs/prds/03-convener/convener-widget.md - Backend:
ConvenerController— innovation spread, learning catalyst - Action: Innovation spread visualization (Rogers diffusion S-curve), learning recommendation cards, champion discovery with skill matching.
- Backend: C#
WidgetRegistry,PluginOrchestrator,MarketplaceEntrymodels - Action: Widget marketplace page: browse, install/uninstall, version management, security sandbox info badges, dependency resolution display.
- PRD:
docs/prds/08-organizational-transformation/org-mesh-widget.md - Action: Organization-level cognitive mesh visualization — department network graph (D3 force layout), capability heatmap, transformation progress tracker with milestones.
- Files:
src/UILayer/web/src/app/— new route directories - Action: Create route directories:
/dashboard,/settings,/agents,/compliance,/analytics,/marketplace. Addloading.tsx,error.tsx,layout.tsxper route group. Implement parallel route loading.
- Files:
src/UILayer/web/src/components/Navigation/ - Action: Sidebar with collapsible sections, breadcrumbs, mobile hamburger menu. Active route highlighting. Responsive drawer (< 768px). Keyboard navigation (arrow keys).
- Files:
src/UILayer/web/src/lib/auth/permissions.ts - Action: Permission map from JWT claims.
<RequireRole>wrapper component. Admin-only routes. Permission-gated buttons/forms. Roles: Admin, Analyst, Viewer. - Depends on: FE-004 ✓
- File:
src/UILayer/web/Dockerfile(new) - Action: Multi-stage build: Node 22 → nginx:alpine. SPA routing config, health check endpoint, runtime env injection via
entrypoint.sh.
- File:
docker-compose.yml(update) - Action: Add
webservice, port 3000,depends_on: [api], volume mount for config.
- File:
.github/workflows/deploy.yml(update) - Action: Build → push to ACR → deploy to Azure Static Web Apps or AKS. Staging + production environments.
- Files:
k8s/base/frontend-deployment.yaml,k8s/base/frontend-service.yaml - Action: Deployment, Service, Ingress with TLS, ConfigMap for
NEXT_PUBLIC_API_BASE_URL, HPA (2–10 replicas).
- File:
infra/modules/frontend/(new) - Action: Azure Static Web Apps module, custom domain, TLS cert, WAF rules, CDN integration.
- Files:
src/UILayer/web/src/components/**/*.test.tsx - Action: Tests for all 47+ components. Jest + Testing Library. Mock API calls with MSW. Snapshot tests for static components.
- Files:
src/UILayer/web/src/lib/api/__tests__/ - Action: Test generated API client against MSW mock server. Verify all 13 controller integrations. Error scenario coverage (401, 403, 500, timeout).
- Files:
cypress/e2e/orplaywright/ - Action: Playwright or Cypress. Flows: login → dashboard → agent management → settings → logout. Cross-browser (Chrome, Firefox, Safari).
- Files:
.storybook/,chromatic.config.js - Action: Chromatic or Percy integration with Storybook. Baseline screenshots for all widget states. PR comments for visual diffs.
- File:
.github/workflows/lighthouse.yml(new) - Action: Lighthouse CI on PRs. Thresholds: Performance >= 80, Accessibility >= 95, Best Practices >= 90, SEO >= 80. Block merge on regression.
- Action: html2canvas + jsPDF. Export current dashboard layout with all widget states. Include timestamp and user info.
- Action: Global keyboard shortcut. Search: pages, agents, widgets, settings, recent actions. Fuzzy matching. Keyboard-navigable results.
- Action: SignalR presence tracking. Avatar bubbles showing active users per dashboard. Cursor/selection sharing (optional).
- Depends on: FE-003
- Action: es-ES, ja-JP, zh-CN (170+ keys each). RTL consideration for future ar-SA. Locale-aware number/date formatting.
- Action: Web app manifest, install prompt, push notification subscription, offline dashboard with cached last-known state.
Each ticket produces a weighted decision matrix and recommendation. Criteria weights: Feature fit 30%, Cost 20%, Integration effort 20%, MCP availability 15%, Community/support 15%. Score 1–5 per candidate.
- Candidates: GitHub Actions (current), Azure DevOps Pipelines, Octopus Deploy
- Evaluate: Multi-environment support, approval gates, rollback capability, audit trail, MCP server availability for AI-driven deployment automation
- Decision matrix: Feature fit (multi-env, blue-green, canary), Cost (per-seat vs per-pipeline), Integration effort (existing GitHub ecosystem), MCP (agent-triggered deploys), Community (docs, extensions)
- Candidates: SonarQube (current), Codacy, CodeClimate, Qodana (JetBrains)
- Evaluate: .NET + TypeScript language coverage, PR integration quality, custom rule authoring, MCP server for AI-agent quality queries
- Decision matrix: Feature fit (language support, rule depth), Cost (OSS vs cloud), Integration effort (CI hooks, IDE plugins), MCP (query findings from agents), Community (plugin ecosystem)
- Candidates: Dependabot (current), Snyk, Sonatype (Nexus Lifecycle), Mend (WhiteSource)
- Evaluate: CVE database coverage, license compliance scanning, auto-remediation PRs, MCP server for agent-driven vulnerability triage
- Decision matrix: Feature fit (CVE coverage, license audit), Cost (free tier limits), Integration effort (GitHub native vs external), MCP (auto-triage from agents), Community (enterprise adoption)
- Candidates: Neo4j, ScapeGraph, Azure AI Search (Cognitive Search)
- Evaluate: Knowledge extraction from codebase, semantic reasoning integration, graph query performance, MCP server for agent knowledge queries
- Decision matrix: Feature fit (graph modeling, semantic search), Cost (hosted vs self-managed), Integration effort (.NET SDK, REST API), MCP (agent-queryable knowledge base), Community (AI/ML ecosystem)
- Candidates: Azure Monitor (current partial), OpenTelemetry + Grafana, Datadog
- Evaluate: .NET auto-instrumentation, distributed tracing, custom dashboards, alerting, MCP server for agent-driven observability queries
- Decision matrix: Feature fit (traces, metrics, logs unified), Cost (per-host, per-GB), Integration effort (.NET SDK, collector setup), MCP (query metrics from agents), Community (CNCF backing, docs)
- Candidates: Azure Boards (current partial), Linear, Serena
- Evaluate: AI agent integration for automated task management, MCP server availability, sprint/kanban support, automation rules
- Decision matrix: Feature fit (automation, API richness), Cost (per-seat), Integration effort (GitHub sync, webhook), MCP (agent-managed backlogs), Community (developer experience)
| Priority | Total | Done | Remaining |
|---|---|---|---|
| P0-CRITICAL (frontend) | 4 | 4 | 0 |
| P1-HIGH (frontend infra) | 6 | 6 | 0 |
| P1-HIGH (UI library — 14b) | 4 | 4 | 0 |
| P1-HIGH (widget PRDs) | 5 | 0 | 5 |
| P2-MEDIUM (widgets + nav) | 8 | 4 | 4 |
| P2-MEDIUM (security) | 1 | 0 | 1 |
| P2-MEDIUM (CI/CD) | 5 | 0 | 5 |
| P2-MEDIUM (testing) | 5 | 0 | 5 |
| P3-LOW (features) | 5 | 0 | 5 |
| DEVOPS (evaluation) | 6 | 0 | 6 |
| Total remaining | 49 | 18 | 31 |
Items: FE-001, FE-004, FE-006, FECICD-001 Status: Complete. All PR comments addressed. CI green.
| Item | Description | Status |
|---|---|---|
| FE-001 | TypeScript API client (openapi-typescript + openapi-fetch) | ✓ Done |
| FE-004 | Auth flow: login page, JWT context, protected routes, middleware | ✓ Done |
| FE-006 | Error handling: ErrorBoundary, Toast, API interceptors | ✓ Done |
| FECICD-001 | Frontend CI job (lint, tsc, jest, coverage) | ✓ Done |
| Blocker | AdaptiveBalance + NIST controller route decorators | ✓ Done |
Also completed: PR review fixes — NotImplementedException stubs replaced in AgentRegistryService (10 methods) and AuthorityService (17 methods), fire-and-forget safety, Guid.Empty fix, SSR-safe ApiBootstrap, middleware JWT validation, Toast memory leak fix, returnTo flow, dependency cleanup.
- All CI checks green (frontend job, CodeQL, .NET build)
- All P1 PR review comments resolved
- CodeQL findings addressed (structured logging — safe pattern)
- package-lock.json committed, dependency issues fixed
Items: CICD-REN-001, CICD-REN-002, CICD-REN-003 Goal: Automate dependency management — automerge minor/patch, auto-assign Codex for major migrations.
| Item | Description | Key Work |
|---|---|---|
| CICD-REN-001 | Renovate automerge for minor/patch | Extend existing patch-only rule to minor+patch. Add platformAutomerge: true for GitHub-native merge. Covers both NuGet and npm deps. |
| CICD-REN-002 | Codex auto-assignment for majors | Add assignees: ["codex[bot]"] rule for major updates. Gate via dependencyDashboardApproval. Codex reads changelog, updates call sites, fixes breaking changes. |
| CICD-REN-003 | agentkit-forge template update | Create issue on agentkit-forge to include Renovate+Codex pattern in bootstrap template for all PhoenixVC repos. |
Prerequisites:
- Enable Codex coding agent in Copilot settings for target repos
- Verify Codex bot username after first assignment (
gh api repos/{owner}/{repo}/pulls/{n} --jq '.assignees[].login') - No branch protection on main (verified — automerge will work without bypass config)
- Renovate minor/patch PRs auto-merging after CI passes
- Major version PRs assigned to Codex, Codex producing migration diffs
- agentkit-forge ticket created for template propagation
Items: FE-002, FE-003, FE-005, FE-007, FE-021, FE-022 Status: Complete. All 6 core items implemented.
| Item | Description | Status |
|---|---|---|
| FE-002 | Replace mocked API with real backend | ✓ Done — DashboardAPI singleton + useDashboardData hook deleted. Root / redirects to /dashboard. Stores fetch from real backend. |
| FE-003 | SignalR real-time client | ✓ Done — useSignalR hook with @microsoft/signalr@10.0.0. HubConnection to /hubs/cognitive-mesh, exponential backoff (1s→30s), ConnectionIndicator in TopBar. |
| FE-005 | Zustand state management | ✓ Done — 5 stores: useAuthStore, useAgentStore, useDashboardStore, useNotificationStore, usePreferencesStore. Persistence middleware for preferences. |
| FE-007 | Loading states + skeletons | ✓ Done — Skeleton, SkeletonCard, SkeletonTable, SkeletonMetric, SkeletonDashboard. loading.tsx and error.tsx per route group. |
| FE-021 | Multi-page routing | ✓ Done — (app) route group with 6 routes: /dashboard, /agents, /settings, /analytics, /compliance, /marketplace. Shared layout with ProtectedRoute. |
| FE-022 | Navigation component | ✓ Done — Sidebar (collapsible), TopBar (breadcrumbs + notification bell), MobileMenu (responsive drawer), Breadcrumbs, ConnectionIndicator. Active route highlighting via usePathname(). |
Deferred to later phases (require backend work):
- httpOnly cookie for refresh token (requires backend
/api/auth/refreshset-cookie endpoint) - Full JWT validation in middleware (requires JWKS endpoint or shared secret config)
- Backend auth middleware in
Program.cs(AddAuthentication/AddAuthorization)
- DashboardAPI mock removed, stores fetching from real backend
- SignalR hook created with reconnection logic and connection indicator
- Navigation between all 6 routes works with loading/error states
- Zustand stores hydrating from API on mount
- Storybook stories for skeleton and navigation components (deferred to Phase 15)
Items: FEUI-001, FEUI-002, FEUI-003, FEUI-004 Status: Complete. All 169 files from CognitiveMeshUI accounted for (153 migrated, 4 removed dead code, 3 superseded, 7 IDE configs, 2 deduplicated).
| Item | Description | Status |
|---|---|---|
| FEUI-001 | Fix or replace broken shadcn/ui components | ✓ Done — All 48 shadcn/ui components working with 27 @radix-ui/* packages installed. Zero TS errors. |
| FEUI-002 | Import design tokens | ✓ Done — tokens/ directory imported (colors, typography, spacing, dimensions, text, object-values). Style Dictionary v5 generates build/css/_variables.css. Imported in globals.css. |
| FEUI-003 | Wire Storybook | ✓ Done — Storybook v10 config (core + react-webpack5 + addon-links). 4 stories migrated. Addon alignment deferred (essentials bundled in v10 core). |
| FEUI-004 | Clean up duplicate/dead component code | ✓ Done — Removed BridgeHeader, FXModePanel, LayoutToolsPanel, VoiceFeedback (dead code). Deduplicated hooks (use-mobile, use-toast). Consolidated type definitions. |
Additional work completed:
- Tailwind CSS v3 → v4 migration (
@tailwindcss/postcss+ CSS-first@config) - Next.js 16 SSR hardening (Suspense boundaries,
typeof windowguards, env fallbacks) - Duplicate route conflict resolved (
/settingsroot deleted, kept(app)/settings)
-
npx tsc --noEmitpasses with zero errors -
Design tokens generating CSS custom properties
-
[~] Storybook running locally (config aligned to v10, addon version alignment deferred)
-
No duplicate type definitions across stores/components
Items: FE-011–FE-015, FE-008, FE-009, FE-010 Goal: Implement the 5 PRD-defined widgets and user-facing settings/profile pages.
| Item | Description | Key Work |
|---|---|---|
| FE-011 | NIST Compliance Dashboard | Maturity score gauge (D3 radial), pillar breakdown cards, evidence upload, gap analysis table, roadmap timeline. 7 endpoints. |
| FE-012 | Adaptive Balance Widget | Interactive spectrum sliders (5 dims), real-time via SignalR, override approval flow, audit trail, recommendations. |
| FE-013 | Value Generation Widget | Upgrade TwoHundredDollarTestWidget + ValueDiagnosticDashboard. Consent flow, radar chart, org blindness heatmap. |
| FE-014 | Impact Metrics Widget | Psychological safety gauge (6 dims), mission alignment radar, adoption timeline, resistance cards. 8 endpoints. |
| FE-015 | Cognitive Sandwich Widget | Phase stepper (Human→AI→Human), HITL approval with diff viewer, cognitive debt burndown, audit log. SignalR. |
| FE-008 | Settings Page | Theme (light/dark/system), language, accessibility (reduced motion, high contrast, font size), privacy consent toggles. |
| FE-009 | Notification Preferences | Channel toggles (email, push, SMS, in-app), category filters, quiet hours with timezone picker. |
| FE-010 | User Profile | Account info, role badges, GDPR consent management, data export, session history. |
- All 5 PRD widgets rendering with real API data
- Each widget has at least 1 Storybook story and 1 unit test
- Settings page persists preferences and applies them (theme, language)
- GDPR consent flow works end-to-end (consent → backend → audit record)
Items: FE-016–FE-020, FE-023, FECICD-002–006 Goal: Additional widgets, role-based access, containerization, and deployment infrastructure.
| Item | Description | Key Work |
|---|---|---|
| FE-016 | Context Engineering Widget | Token budget visualizer, frame composition editor, context window gauge. |
| FE-017 | Agentic System Control | Upgrade AgentControlCenter. Lifecycle CRUD, authority scope config, orchestration trigger with streaming results. |
| FE-018 | Convener Widget | Rogers diffusion S-curve, learning recommendations, champion discovery with skill matching. |
| FE-019 | Widget Marketplace | Browse/install/uninstall, version management, security sandbox badges, dependency resolution. |
| FE-020 | Organizational Mesh Widget | Department network graph (D3 force), capability heatmap, transformation progress tracker. |
| FE-023 | Role-Based UI Gating | <RequireRole> wrapper, admin-only routes, permission-gated buttons. Roles: Admin, Analyst, Viewer. |
| FECICD-002 | Frontend Docker | Multi-stage Node 22 → nginx:alpine. SPA routing, health check, runtime env injection. |
| FECICD-003 | docker-compose integration | web service, port 3000, depends_on api, config volume. |
| FECICD-004 | Deploy pipeline | Build → ACR → Azure Static Web Apps. Staging + production. |
| FECICD-005 | K8s manifests | Deployment, Service, Ingress with TLS, ConfigMap, HPA (2–10 replicas). |
| FECICD-006 | Terraform frontend | Azure Static Web Apps module, custom domain, TLS, WAF, CDN. |
- Docker build succeeds and container starts cleanly
-
docker-compose upbrings up full stack (API + frontend + dependencies) - RBAC enforced: Viewer cannot access admin routes, roles shown in UI
- Staging deployment works via CI pipeline
- All widgets render (even if some show placeholder data)
Items: FETEST-001–005, FE-024–FE-028 Goal: Comprehensive test coverage, visual regression, performance monitoring, and advanced features.
| Item | Description | Key Work |
|---|---|---|
| FETEST-001 | Component unit tests (80%) | All 47+ components. Jest + Testing Library. MSW mocks. Snapshot tests for static components. |
| FETEST-002 | API integration tests | Generated client vs MSW mock server. All 13 controllers. Error scenarios (401, 403, 500, timeout). |
| FETEST-003 | E2E tests | Playwright or Cypress. Login → dashboard → agents → settings → logout. Cross-browser. |
| FETEST-004 | Visual regression | Chromatic or Percy + Storybook. Baseline screenshots. PR comments for visual diffs. |
| FETEST-005 | Lighthouse CI | Performance >= 80, A11y >= 95, Best Practices >= 90, SEO >= 80. Block merge on regression. |
| FE-024 | Dashboard export | html2canvas + jsPDF. PDF/PNG export with timestamp and user info. |
| FE-025 | Command palette (Cmd+K) | Global search: pages, agents, widgets, settings, recent actions. Fuzzy matching. |
| FE-026 | Collaboration presence | SignalR presence. Active user avatars per dashboard. Optional cursor sharing. |
| FE-027 | Additional locales | es-ES, ja-JP, zh-CN (170+ keys each). RTL prep for ar-SA. Locale-aware formatting. |
| FE-028 | PWA enhancements | Web manifest, install prompt, push notifications, offline dashboard (cached last state). |
- Unit test coverage >= 80% across all components
- All E2E flows passing in CI
- Lighthouse scores meeting thresholds
- No visual regressions from baseline
- PWA installable and functional offline (read-only)
Items: DEVOPS-001–006 Goal: Evaluate and select best-fit DevOps tools. Each ticket produces a weighted decision matrix and recommendation.
| Ticket | Category | Candidates | MCP Integration |
|---|---|---|---|
| DEVOPS-001 | Deployment Pipeline | GitHub Actions, Azure DevOps, Octopus Deploy | Agent-triggered deploys, auto-rollback |
| DEVOPS-002 | Code Quality | SonarQube, Codacy, CodeClimate, Qodana | Agent-queryable findings, auto-fix suggestions |
| DEVOPS-003 | Dependency Security | Dependabot, Snyk, Sonatype, Mend | Agent-driven vulnerability triage |
| DEVOPS-004 | AI Knowledge Graphs | Neo4j, ScapeGraph, Azure AI Search | Agent-queryable knowledge base |
| DEVOPS-005 | Observability | Azure Monitor, OpenTelemetry+Grafana, Datadog | Agent-driven metrics queries, auto-alerting |
| DEVOPS-006 | Task Orchestration | Azure Boards, Linear, Serena | Agent-managed backlogs, auto-triage |
Weighted criteria: Feature fit 30%, Cost 20%, Integration effort 20%, MCP availability 15%, Community 15%
- Decision matrix reviewed by team for each category
- Budget approved for selected tools
- PoC completed for top pick in each category
- MCP server built or validated for at least 3 of 6 categories
Updated: 2026-03-11 | Backend 100% complete (70/70). Frontend Phase 13 ✓, Phase 14 ✓ (6/6), Phase 14b ✓ (4/4). Phase 15 next (Widgets & User Settings). Remaining: 31 items across Phases 15–18.