🏗️ Architecture: Should Vitto directly connect to other MongoDB databases or ...

Date: 2026-02-11
Time: 19:39:10 UTC
Session: gspace1-106.3
Agent: backend
Confidence: 92%
Status: ✅ Documented

📋 Decision Context

Question:

Should Vitto directly connect to other MongoDB databases or go through APIs?

Task: Session 106: Vitto cross-database knowledge federation architecture

🎯 Decision Made

Chosen:

API-first with a Knowledge Sync Pipeline. Vitto should NOT directly connect to gograbspace MongoDB. Instead: (1) backend-ai-api exposes a /knowledge/feed endpoint that packages relevant knowledge (decisions, patterns, conventions) in a Vitto-digestible format, (2) Vitto periodically syncs via a scheduled job (KnowledgeSyncService), (3) Vitto writes its execution outcomes back via the existing emitOrchestrationEvent and logDecision APIs. This preserves service boundaries while enabling bidirectional learning.

Options Considered:

• No alternatives documented

💭 Reasoning

Direct MongoDB cross-connection violates service boundaries and creates tight coupling. If gograbspace schema changes, Vitto breaks. API-mediated sync is slower but maintainable. The BackendAiClientService already has methods like findSimilarDecisions, getConventions, getLearningContext. The gap is that these are called on-demand per task, not systematically ingested into Vitto's Mind. A scheduled sync job solves this without architectural coupling.

📁 Files Involved

• apis/backend-vitto-api/src/app/connectivity/backend-ai-client.service.ts
• apis/backend-vitto-api/src/app/mind/mind.service.ts

🔗 Related

• Decision ID: dev_1770838749950_xq5kzbg
• Session: 106
• Workspace: gspace1

Auto-generated by MCP Smart Consolidation