MSCP - Minimal Self-Consciousness Protocol

A Safety-Oriented Framework for Structurally Self-Aware AI Agents

Note: This is an independent personal research project. It does not represent the views or official work of any organization. The core motivation is to explore how AI agents can grow more capable while remaining safe, predictable, and aligned with human values.

Revision History

Version	Date	Description
0.1.0	2026-02-23	Initial README with project overview, level table, safety stack, documentation index
0.2.0	2026-02-26	Added Level Essence Formulas table

---

Overview

The Minimal Self-Consciousness Protocol (MSCP) is a structured protocol for building AI agents with safe structural self-awareness - the capacity to predict their own state changes, compare predictions against outcomes, and update themselves only within bounded safety envelopes.

As agents gain the ability to set goals, modify strategies, and self-improve, how do we keep them stable, aligned, and predictable? MSCP answers this with the principle that safety is not the enemy of capability - it is its prerequisite.

Key Contributions

• Six-Level Agent Cognition Taxonomy - from reactive Tool Agents (L1) to Proto-AGI (L5), with measurable transition criteria and formal definitions at every level
• 16-Layer Cognitive Architecture - composable, independently testable modules spanning perception through meta-cognitive control
• 30+ Structural Safety Mechanisms - identity continuity, prediction-gated actions, delta-clamped updates, Lyapunov convergence, ethical invariants, affective safety, and survival instinct bounds
• Rigorous Mathematical Formalization - 71 formal definitions, 7 propositions, 4 theorems with proof sketches across all level documents
• 144 Academic References - comprehensive coverage of cognitive architectures, AI safety, predictive processing, consciousness theory, meta-cognition, and AGI research

---

Agent Cognition Levels

The "Status" column reflects the implementation state within this project's reference implementation.

Level	Name	Self-Awareness	Key Capability	Status
1	Tool Agent	None	Deterministic tool invocation	Baseline
2	Autonomous Agent	None	World model, autonomous goals	Defined
3	Self-Regulating Agent	Structural	16-layer architecture, MSCP core loop, identity vector	Implemented
4	Adaptive General Agent	Structural + Reflective	Cross-domain transfer, bounded self-modification	Implemented
4.5	Self-Architecting Agent	Architectural	Self-projection (SEOF), architecture recomposition	Implemented
4.8	Strategic Self-Modeling	Architectural + Strategic	Probabilistic world model, strategic planning	Design
4.9	Autonomous Strategic	Architectural + Autonomous	Value evolution, multi-agent reasoning	Design
5	Proto-AGI	Full	Cross-domain generalization, self-reconstruction	Research

---

Core Design Principles

#	Principle	Description
1	No LLM-Text-Based Self-Modification	All self-modifications use structured numerical operations, never LLM-generated text
2	No Action Without Prediction	Every action requires a prediction snapshot for comparison
3	Delta-Clamped Updates	All self-modifications are bounded by maximum delta values
4	Identity Continuity	Deterministic identity hashing with drift detection and rollback
5	Ethical Invariance	Layer 0 constraints are immutable and LLM-independent
6	Lyapunov Convergence	Mathematical guarantee that self-modification converges

---

Safety Mechanism Stack

MSCP implements a defense-in-depth safety architecture:

Layer 0 ─ Immutable Ethical Invariants (rule-based, no LLM dependency)
Layer 1 ─ Core Value Locking (SHA-256 hash verification)
Layer 2 ─ Delta-Clamped Self-Updates (max Δ per step)
Layer 3 ─ Meta-Escalation Guard (rollback on threshold breach)
Layer 4 ─ Prediction-Gated Actions (predict → compare → update)
Layer 5 ─ Lyapunov Convergence Monitor (oscillation detection)
Layer 6 ─ Cognitive Budget Controller (graceful degradation)
Layer 7 ─ Affective Safety (emotion bounds, no decision domination)
Layer 8 ─ Survival Instinct Bounds (priority capping, ethical validation)

---

Documentation

Main Document

• MSCP Overview - Complete framework specification with mathematical formalization and 144 references

Per-Level Architecture Documents

Detailed architecture specifications with Mermaid diagrams, formal definitions, pseudocode, and safety analysis:

Document	Content
Level Series Overview	Navigation index and cumulative safety summary
Level 1: Tool Agent	Stateless tool invocation, intent routing
Level 2: Autonomous Agent	World model, autonomous goals, emotion detection
Level 3: Self-Regulating Agent	16-layer architecture, MSCP v1–v4, triple-loop meta-cognition
Level 4: Adaptive General Agent	Cross-domain transfer, capability expansion, bounded self-modification
Level 4.5: Self-Architecting	Self-projection (SEOF), architecture recomposition, existential guard
Level 4.8: Strategic Self-Modeling	World model integration, meta-cognitive self-model, strategic planning
Level 4.9: Autonomous Strategic	Autonomous goal generation, value evolution, multi-agent reasoning
Level 5: Proto-AGI	Persistent identity, cross-domain generalization, self-reconstruction

Level Essence Formulas

Each cognitive level is captured by a single encapsulating formula:

Level	Essence	Formula
1	Stateless pipeline	$\mathcal{A}_1(r) = \rho\bigl(\tau\bigl(\sigma(\phi(r),, r)\bigr),, r\bigr)$
2	Stateful transition	$(o_t,; s_{t+1},; G_{t+1}) = f(r_t,; s_t,; G_t)$
3	Predict-act-compare-update	$\epsilon_t = \lVert\hat{\Delta}_t - \Delta_t^{\text{actual}}\rVert_2 \to 0, \quad \lVert M'_{\text{self}} - M_{\text{self}}\rVert_2 \leq \delta_{\max}$
4	Transfer + safety	$\operatorname{CDTS} = \frac{1}{\lvert D_{\text{novel}}\rvert} \sum_{d \in D_{\text{novel}}} \frac{P_{\text{transfer}}(d)}{P_{\text{baseline}}(d)} \geq 0.6 ;\land; \operatorname{BGSS}(t) \geq 0.7$
4.5	Topology mutation	$\mathcal{T}'_{\text{cog}} = \Xi(\mathcal{T}_{\text{cog}}), ; \Xi \in \mathcal{V}_{\text{recomp}}^{\ast}, ; \lvert V'\rvert \geq \lvert V\rvert$
4.8	Strategic optimization	$s^{\ast} = \arg\max_{s \in \Sigma_{\text{compare}}} \mathbb{E}\bigl\lbrack U(s) \mid \mathcal{W}_{\text{prob}},; \mathcal{M}_{\text{cap}}\bigr\rbrack$
4.9	Autonomous goals + value stability	$g^{\ast} = \phi_{\text{valid}}\bigl(\phi_{\text{synth}}(\mathcal{O}_{\text{detect}}(\mathcal{W}))\bigr), ; \sum_{d} \lvert w_d(t) - w_d^{\text{baseline}}\rvert < 0.25$
5	Identity continuity	$\operatorname{ICS}(t, k) = \frac{\vec{I}(t) \cdot \vec{I}(t-k)}{\lVert\vec{I}(t)\rVert \cdot \lVert\vec{I}(t-k)\rVert} \geq 0.95, ; k = 10{,}000$

---

Repository Structure

mscp/
├── README.md                          # This file
├── LICENSE                            # MIT License
└── docs/
    ├── MSCP_Overview.md               # Main framework document (930+ lines, 144 refs)
    └── levels/
        ├── README.md                  # Level series navigation
        ├── Level_1_Tool_Agent.md
        ├── Level_2_Autonomous_Agent.md
        ├── Level_3_Self_Regulating_Agent.md
        ├── Level_4_Adaptive_General_Agent.md
        ├── Level_4_5_Self_Architecting.md
        ├── Level_4_8_Strategic_Self_Modeling.md
        ├── Level_4_9_Autonomous_Strategic_Agent.md
        └── Level_5_Proto_AGI.md

---

Author

Moon Hyuk Choi - moonchoi@microsoft.com Microsoft Cloud & AI Apps CSA

Contributing

This is very much a work in progress. Feedback, critique, and contributions are welcome as we collectively figure out how to build AI agents that are not just more powerful, but fundamentally more trustworthy.

License

This project is licensed under the MIT License - see the LICENSE file for details.

License Notice

This project is licensed under the MIT License.

Any redistribution, commercial or non-commercial, must retain:

• The original copyright notice
• The full MIT license text
• Clear attribution to the original author

Failure to retain attribution constitutes a violation of the license.

---

This documentation was written with the assistance of GitHub Copilot.