§ 4 - Core Mechanisms
Mechanism 03 · Intentional StatelessnessFresh Instance Protocol
Every cognitive cycle launches a completely new LLM instance receiving exactly three inputs: the Turn Goal, the Regulations, and verified Memory Facts. No accumulated conversation history enters.
This is not a missing feature - it is a deliberate architectural choice grounded in Baddeley's working-memory model: residue from a previous task acts as interference on a new one. Fresh Instance keeps per-cycle input size linear, makes cost predictable, and structurally blocks confirmation bias.
Architectural Advantages
- •Linear Context Costs - Prevents ballooning tokens due to long chat threads.
- •Eliminates Goal Drift - Keeps the model focused only on the current turn goals and verified memory state.
- •Zero Context Leaks - Ensures previous tasks cannot bleed context or contaminate current reasoning.
Examine Governance Alignment
Explore how the SCL architecture aligns with the EU AI Act requirements.