Research & Publications

SCD Protocol v3.1

PUBLISHED

Structured Contextual Distillation — The governance layer that enforces truth-state handling, memory rights, and continuity guarantees in LLM systems.

The SCD Protocol defines how AI systems should classify claims (FACT/ESTIMATE/UNKNOWN), manage memory operations, and maintain cryptographic anchors for drift detection. It's the core specification that makes MirrorDNA systems trustworthy.

ActiveMirrorOS White Paper

PUBLISHED

Wrapper-First AI Systems for Governed Intelligence — Academic documentation of the wrapper-first architecture that puts governance before inference.

Traditional AI systems run inference first, then try to filter outputs. ActiveMirrorOS inverts this: governance runs first, validating inputs and classifying intent before any inference occurs. This creates deterministic security guarantees.

ADHD-AI Interface Discovery

VAULTED

Hypothesis: Neurodivergent cognitive patterns — specifically ADHD traits like low tedium tolerance, non-linear thinking, pattern-jumping, and hyperfocus — may be advantages in AI collaboration rather than limitations.

Traditional programming requires sustained linear focus. AI orchestration rewards rapid context-switching, pattern recognition across domains, and comfort with ambiguity. This hypothesis proposes ADHD as a native interface for AI collaboration.

The Orchestrator Model

WORKING PAPER

Reframe: The shift from CEO (managing humans who execute) to Orchestrator (directing AI systems that execute).

Traditional scaling requires hiring. Each hire adds communication overhead, coordination costs, alignment maintenance. The Orchestrator model scales with compute instead of headcount. One human directs multiple AI systems that maintain coherent execution across domains.

Key insight: What scales with compute > What scales with people.

Active Mirror Framework

PUBLISHED

Purpose: Universal decision-making artefact that transforms AI assistants into structured thinking partners. Not just generate text — engage in structured deliberation.

The framework provides templates for: decision analysis, trade-off evaluation, risk assessment, and action prioritization. It converts open-ended AI conversations into structured output.