AI Routes Cleanup & Organization Plan

🎯 Goals

1. Remove outdated/broken components
2. Connect all components to real backend-ai-api
3. Use flash-persistence for local data
4. Ensure consistent data flow
5. Organize routes logically

📋 Current Routes Analysis

✅ Working Components (Keep & Improve)

1. Training Center - /ai/ai-agents-training-center

• ✅ Connected to ModuleIntelligenceHarvesterService
• ✅ Uses PatternExtractorService
• ✅ Real data from flash-persistence
• 🔧 Action: Ensure backend sync works

2. Agent Knowledge Viewer - /ai/agent-knowledge/:agentId?

• ✅ Connected to backend API (/ai-api/platform/agents/:agentId/knowledge)
• ✅ All services properly wired
• ✅ Fixed takeUntilDestroyed errors
• 🔧 Action: Test with real agent data

3. AI Dashboard - /ai/dashboard

• ✅ Using ModuleIntelligenceHarvesterService
• ✅ Using PatternExtractorService
• 🔧 Action: Remove any remaining mock data calls

⚠️ Needs Update (Fix & Connect)

4. Platform Intelligence - /ai/platform-intelligence

• Component: PlatformIntelligenceDashboardComponent
• 🔧 Action: Connect to /ai-api/platform/agents API
• 🔧 Action: Show real platform agents and their training status

5. Cursor Agents Dashboard - /ai/cursor-agents

• Component: CursorAgentsDashboardComponent
• 🔧 Action: Connect to /ai-api/agents API
• 🔧 Action: Show development agents (cursor-angular-agent, cursor-backend-agent, etc.)

6. Space Assistants - /ai/vitto-hubs

• Component: SpaceAssistantsDashboardComponent
• 🔧 Action: Connect to tenant-specific AI APIs
• 🔧 Action: Show tenant-isolated assistants

7. AI Agents Dashboard - /ai/agents-dashboard

• Component: AiAgentsDashboardComponent
• 🔧 Action: Consolidate with other agent views?
• 🔧 Action: Or make this the "All Agents" overview

8. Agent Mesh - /ai/mesh

• Component: AgentMeshVisualizationComponent
• 🔧 Action: Connect to real agent communication data
• 🔧 Action: Use WebSocket for real-time updates

9. Memory Browser - /ai/memory

• Component: MemoryBrowserComponent
• 🔧 Action: Connect to /ai-api/memory endpoints (if they exist)
• 🔧 Action: Browse flash-persistence data

10. AI Playground - /ai/playground

• Component: AiPlaygroundComponent
• 🔧 Action: Connect to real AI services
• 🔧 Action: Test agent capabilities interactively

11. Visual Memory Map - /ai/memory-map

• Component: VisualMemoryMapComponent
• 🔧 Action: Connect to vector memory data
• 🔧 Action: Visualize knowledge clusters

❌ Outdated/Broken (Remove or Major Refactor)

12. AI Tester - /ai/ai-tester

• Component: AiTesterComponent
• ❌ Problem: Tons of commented-out code, circular dependencies
• ❌ Problem: Trying to manage API keys (should be backend responsibility)
• ❌ Problem: Not using our standard patterns
• 🗑️ Action: REMOVE or completely rewrite as "AI Service Testing Tool"
  • If keeping: Remove all key management UI
  • If keeping: Only test AI responses through backend proxies
  • If keeping: Make it a developer tool, not user-facing

13. AI Guide - /ai/guide

• Component: AiGuideComponent
• ⚠️ Review: Is this still useful or is it outdated documentation?
• 🔧 Action: Update or remove

14. AI Config - /ai/ai-config

• Component: AiConfigComponent
• ⚠️ Review: What configs do users need to set?
• 🔧 Action: Simplify or remove (most config should be backend)

15. AI Integration Examples - /ai/integration-examples

• Component: AiIntegrationExamplesComponent
• ⚠️ Review: Is this developer documentation?
• 🔧 Action: Move to docs or remove

🆕 Missing (Consider Adding)

1. Agent Training History - /ai/training-history

   • Show training sessions, outcomes, improvements over time

2. Pattern Library - /ai/patterns

   • Browse all learned patterns across all agents

3. Real-time Agent Monitor - /ai/monitor

   • Live view of agent activity, tasks, performance

🛠️ Implementation Plan

Phase 1: Remove/Clean (Immediate)

1. ✅ Fix takeUntilDestroyed errors in agent-knowledge
2. 🔄 Remove or drastically simplify ai-tester
3. Remove/consolidate duplicate agent dashboards
4. Remove outdated demo/example components

Phase 2: Connect to Real APIs (High Priority)

1. Update Platform Intelligence Dashboard
2. Update Cursor Agents Dashboard
3. Update Space Assistants Dashboard
4. Connect Memory Browser
5. Connect Agent Mesh

Phase 3: Polish & Enhance (Medium Priority)

1. Add real-time WebSocket updates
2. Add comprehensive error handling
3. Add loading states everywhere
4. Ensure consistent UI/UX

Phase 4: New Features (Low Priority)

1. Add training history viewer
2. Add pattern library browser
3. Add real-time monitoring

📊 Backend API Endpoints Available

Platform AI (/ai-api/platform/agents)

• GET /:agentId/knowledge - Full agent knowledge
• GET /:agentId/patterns - Learned patterns
• GET /:agentId/decisions - Successful decisions
• GET /:agentId/anti-patterns - Failed approaches
• GET /:agentId/specializations - Specialization areas
• POST /:agentId/generate-docs - Generate documentation
• POST /:agentId/import-from-docs - Import from docs

Module Intelligence (/ai-api/ai-agents/module-intelligence)

• POST /harvest - Sync harvested intelligence
• GET /modules - All modules
• GET /modules/:moduleId - Single module
• GET /patterns/:category - Patterns by category
• POST /find-similar - Find similar modules
• POST /outcomes/track - Track outcome
• GET /outcomes/:outcomeId - Get outcome
• GET /outcomes/stats - Outcome statistics
• POST /outcomes/report-to-mcp - Report to MCP
• POST /patterns/extract - Extract patterns
• GET /patterns/stats - Pattern statistics
• POST /docs-data/sync - Sync docs/data

Agent Tasks (/ai-api/agents)

• Task management endpoints (mostly TODOs)

🎨 UI/UX Standards

1. Loading States: All API calls must show loading indicators
2. Error Handling: Graceful fallbacks, user-friendly error messages
3. Empty States: "No data yet" messages with helpful actions
4. Consistent Design: Use Material Design components
5. Real-time Updates: Use signals for reactive UI

🔄 Data Flow

Frontend Components
    ↓
Angular Services (with signals)
    ↓
HttpClient → `/ai-api/*` (proxied to port 3001)
    ↓
backend-ai-api (NestJS)
    ↓
MongoDB (AI databases) + Flash-persistence sync

✅ Success Criteria

1. ❌ No mock data anywhere
2. ✅ All components connect to real APIs or flash-persistence
3. ✅ No circular dependencies or commented-out code
4. ✅ Consistent error handling
5. ✅ Fast, responsive UI with loading states
6. ✅ Clear, logical route organization
7. ✅ Agent knowledge visible and useful

Created: 2025-01-16 Status: In Progress