♻️ Refactor: How should ai-generation-hub's broken generation routing be ...

Date: 2026-03-11
Time: 16:18:19 UTC
Session: gspace1-204.5
Agent: angular
Confidence: 95%
Status: ✅ Documented

📋 Decision Context

Question:

How should ai-generation-hub's broken generation routing be fixed?

Task: Not specified

🎯 Decision Made

Chosen:

Rerouted startGeneration() to delegate to the existing startVittoGeneration() method which uses the correct backend-vitto-api pipeline. Removed mapModeToWorkflowType() and deprecated startStreamingWorkflow(). This means all 3 generation modes (module, galaxy, universe) now go through the same Vitto 6-phase pipeline.

Options Considered:

• No alternatives documented

💭 Reasoning

The service already had startVittoGeneration() (Session 108) that correctly routes through backend-vitto-api. But startGeneration() was calling the broken backend-ai-api workflow instead. Rather than fixing the broken workflow (which would be massive), we reuse the working path. The Vitto pipeline handles all cases: maxModules=1 for module mode, maxModules=10 for galaxy, maxModules=20 for universe.

📁 Files Involved

• libs/libs-ai/ai-generation-hub/src/lib/services/generation-hub.service.ts

🔗 Related

• Decision ID: dev_1773245899880_d7rgmwt
• Session: 204
• Workspace: gspace1

Auto-generated by MCP Smart Consolidation