🏗️ Architecture: What are the critical gaps preventing Vitto from becoming a ...

Date: 2026-03-17
Time: 23:27:30 UTC
Session: gspace1-231.2
Agent: backend
Confidence: 92%
Status: ✅ Documented

📋 Decision Context

Question:

What are the critical gaps preventing Vitto from becoming a Jarvis-level super agent, and what is the implementation roadmap?

Task: Analyze Vitto backend architecture and create comprehensive plan to evolve into Jarvis-level super agent

🎯 Decision Made

Chosen:

6-phase plan: Phase 0 (SSE auth + AgentProcess redirect + audit TTL), Phase 1 (tiered sub-agent tools), Phase 2 (closed-loop skill learning + persona adaptation + cross-channel genome), Phase 3 (OAuth credential delegation for external services), Phase 4 (proactive intelligence with presence awareness), Phase 5 (observability + eval framework)

Options Considered:

1. Single monolithic sprint to implement all gaps at once
2. Focus only on sub-agent capabilities and skip intelligence loops
3. Rebuild from scratch with different architecture
4. 6-phase incremental plan building on existing solid foundation

💭 Reasoning

Deep analysis of 45+ services, 53 MongoDB collections, 17 tool plugins, 60+ frontend components, and 12 design specs revealed that the foundation is strong (4-phase engine, smart routing, memory, knowledge graph, persona, trust escalation all working). Most previously documented gaps (S216, S225.4) have been resolved. The real gaps are: (1) sub-agents are read-only spectators with only 4 tools, (2) no closed-loop skill learning, (3) persona doesn't adapt to user preferences, (4) ~114 @Public() SSE endpoints with no auth, (5) AgentProcessService still bypasses Vitto on ai-api, (6) cross-channel memory is siloed per-session, (7) no OAuth credential delegation for external services, (8) proactive system lacks user presence awareness. Phased approach ensures security first, then capability, then intelligence loops, then external reach.

📁 Files Involved

• APIs/backend-vitto-api/src/app/engine/sub-agent.service.ts
• APIs/backend-vitto-api/src/app/engine/agent-loop.service.ts
• APIs/backend-vitto-api/src/app/engine/phase-executor.service.ts
• APIs/backend-vitto-api/src/app/engine/persona.service.ts
• APIs/backend-vitto-api/src/app/engine/skill-extraction.service.ts
• APIs/backend-vitto-api/src/app/mind/mind.service.ts
• APIs/backend-vitto-api/src/app/engine/vitto-engine.service.ts
• APIs/backend-vitto-api/src/app/connectivity/connectivity.module.ts

🔗 Related

• Decision ID: dev_1773790050109_3vtmhc2
• Session: 231
• Workspace: gspace1

Auto-generated by MCP Smart Consolidation