Superseded by Neo-V3. Archived for history.

NEO-LAB Control Plane

Deterministic Routing and Governance Layer

This repository preserves the control plane layer of NEO-LAB: the deterministic controller (neo_loop.ps1) and chat client (neo_chat.ps1) that classify intent, load explicit state and personas, route each request to exactly one model, and enforce the output contracts. It sits on top of the atomic file-based queue substrate and is where routing rules, modes, and governance constraints live.

NEO-LAB as a whole is a local, deterministic cognitive control plane for orchestrating multiple locally hosted large language models (LLMs) as role-separated inference engines under explicit governance, bounded memory, and fully inspectable state.

All execution is local. All state is explicit. All behavior is deterministic.

---

System Overview

NEO-LAB implements a control-plane architecture for AI inference rather than a traditional chatbot interface. It treats LLMs as stateless, interchangeable compute engines governed by a stateful, deterministic controller.

The system prioritizes:

• Predictability over emergence
• Inspectability over opacity
• Governance over autonomy

This makes NEO-LAB suitable for professional, technical, and safety-conscious workflows.

---

What NEO-LAB Is

NEO-LAB is:

• A deterministic router (rule-based intent → model selection)
• A governed control plane (explicit state, modes, and contracts)
• A file-driven IPC system (atomic, concurrency-safe)
• A local-first AI system (offline-capable, no cloud dependency)
• A single-response generator for summaries, detailed explanations, and multi-page reports

NEO-LAB is not:

• An autonomous agent
• A self-modifying system
• A tool-execution framework
• A blended or ensemble model
• A cloud service

---

Core Architecture (Atomic File-Based IPC)

NEO-LAB uses an atomic per-message file queue to eliminate race conditions and ensure deterministic execution.

queue_v2/
├─ inbox/                  # One JSON file per user message
├─ outbox/
│  └─ <message_id>/
│     ├─ status.json       # Execution phase, progress counters
│     └─ response.txt      # Streaming model output
└─ processed/              # Archived input messages

Processing Flow

User
 ↓
neo_chat.ps1
 ↓  (atomic JSON message)
queue_v2/inbox/<id>.json
 ↓
neo_loop.ps1
 ├─ intent classification
 ├─ explicit state & persona loading
 ├─ deterministic model routing
 ├─ streaming inference
 └─ bounded memory update
 ↓
queue_v2/outbox/<id>/response.txt

Key properties:

• No message overwrites
• Safe concurrency
• Replayable execution
• Fully inspectable artifacts

---

Model Roles (One Active Model per Request)

NEO-LAB routes requests deterministically. Exactly one model is active per request.

Typical role mapping (configurable):

Cognitive Role	Model
Chat / Persona	dolphin-llama3
Code	deepseek-coder-v2
Analysis	deepseek-r1
Vision	qwen2.5-vl

Model availability is queried dynamically via Ollama. If a target model is unavailable, NEO-LAB fails closed or falls back safely.

---

Output Modes (Single-Response Guarantees)

NEO-LAB is designed to produce one complete response per request, regardless of domain.

Supported one-shot modes:

• /summary <topic> — concise, complete response
• /detail <topic> — detailed technical explanation
• /report <topic> — single multi-page structured report

Report Writer Mode

When enabled (/reportmode on), NEO-LAB enforces a strict output contract.

Required sections:

1. Executive Summary
2. Scope & Assumptions
3. Core Analysis
4. Methods / Models / Math (if applicable)
5. Implementation (if applicable)
6. Risks, Limitations, Verification Checklist
7. References / Source Guidance (if applicable)

No follow-up questions. No partial answers. One complete professional deliverable.

---

Streaming Output & Progress Visibility

NEO-LAB streams output incrementally while models generate.

Progress is written to:

queue_v2/outbox/<id>/status.json

Including:

• Current execution phase
• Characters written
• Streaming activity indicators

This ensures transparency during long-running analyses and reports.

---

Memory System (Bounded & Inspectable)

Memory is:

• JSON-based
• Explicitly bounded
• Stored on disk
• Separated by intent

queue/
├─ memory_chat.json
├─ memory_code.json
├─ memory_analysis.json
└─ memory_vision.json

There are:

• No embeddings
• No hidden vectors
• No implicit recall

Memory can be inspected or cleared at any time.

---

Governance & Safety Posture

NEO-LAB is designed with explicit safety constraints:

• No autonomy
• No self-execution
• No self-modification
• No privilege escalation
• Human remains root authority

The system favors control, auditability, and predictability over emergent behavior.

---

Requirements

• Windows 10 / 11
• PowerShell 5.1
• Ollama (local inference server)

Recommended Ollama models:

• dolphin-llama3
• deepseek-coder-v2
• deepseek-r1
• qwen2.5-vl

---

Quick Start

Start the control loop (Terminal A):

cd <repo-root>
.\neo_loop.ps1

Start the chat client (Terminal B):

cd <repo-root>
.\neo_chat.ps1

Example:

/report Write a multi-page engineering report explaining why atomic IPC queues prevent race conditions.

QA Harness (Optional)

A simple QA harness validates report structure and completeness:

powershell -ExecutionPolicy Bypass -File tests\qa\qa_harness.ps1

Results are written to tests/qa/qa_results.json.

License

See LICENSE.

Disclaimer

NEO-LAB provides general informational output only. For legal, medical, or financial decisions, consult qualified professionals and primary sources.