Architecture: How should mastery level control Vitto's behavior during cod...

Documentation for Sessions

🏗️ Architecture: How should mastery level control Vitto's behavior during cod...

Date: 2026-03-10
Time: 01:43:47 UTC
Session: gspace1-196.3
Agent: angular
Confidence: 88%
Status: ✅ Documented

📋 Decision Context

Question:

How should mastery level control Vitto's behavior during code generation?

Task: Not specified

🎯 Decision Made

Chosen:

Structured prompt injection rather than direct autonomy service integration. Mastery context is injected as behavioral instructions in the system prompt: intern='explain everything, ask for confirmation', principal='full autonomy, report outcomes only'. This lets the LLM adapt continuously rather than only at checkpoint boundaries.

Options Considered:

Direct AutonomyService integration (circular dep risk, checkpoint-only)
Middleware-based behavior adjustment (over-engineered for current needs)
Prompt injection with structured behavioral directives (chosen)

💭 Reasoning

Direct AutonomyService integration would create circular dependencies (UsersModule → MindModule → EngineModule → UsersModule) and only controls behavior at checkpoint boundaries. Prompt-based behavioral control is: (1) architecturally clean (no circular deps), (2) continuously adaptive (not just at checkpoints), (3) simpler to implement and debug, (4) naturally composable with other context. The structured prompt includes BEHAVIOR directive, WEAK AREAS, STRONG AREAS, AVOID pitfalls, and RECENT IMPROVEMENT — giving the LLM actionable behavioral guidance.

📁 Files Involved

APIs/backend-vitto-api/src/app/users/job-mastery.service.ts
APIs/backend-vitto-api/src/app/module-generation/context/generation-context.service.ts

🔗 Related

Decision ID: dev_1773107027404_txkzrv1
Session: 196
Workspace: gspace1

Auto-generated by MCP Smart Consolidation