neutron-os 0.9.0

Nuclear facility digital operations platform — extends Axiom

Neutron OS

Nuclear Energy Unified Technology for Research, Operations & Networks

A modular digital platform for nuclear facilities that unifies data management, operations tracking, experiment scheduling, and analytics — replacing fragmented workflows (paper logs, spreadsheets, phone calls) with integrated digital tools.

Key Capabilities

• Signal Pipeline — Ingest signals from voice memos, Teams, GitLab, Linear, and freetext; extract structured insights; publish weekly briefings
• Publisher — Document lifecycle management with provider-based generation, storage, and review
• Interactive Agent — Chat-based assistant with facility context (neut chat)
• Self-Diagnostics — AI-powered troubleshooting (neut doctor)
• Extension System — Scaffold and manage facility-specific extensions (neut ext)
• MCP Server — IDE integration via Model Context Protocol (neut serve-mcp)

Quick Start

# Clone
git clone https://github.com/benjaminbooth/neutron-os-core.git
cd neutron-os-core
pip install -e ".[all]"

# Or use the bootstrap script (creates venv, installs, sets up direnv)
./scripts/bootstrap.sh

# Verify
neut --help
neut doctor

# Run the onboarding demo
neut demo run collaborator

See CONTRIBUTING.md for detailed setup instructions including direnv, venv, and troubleshooting.

Repository Structure

Neutron_OS/
├── src/neutron_os/       # Python package root
│   ├── neut_cli.py       #   CLI entry point
│   ├── infra/            #   Shared infra (gateway, orchestrator, auth)
│   └── extensions/       #   Extension system
│       └── builtins/     #     Builtin extensions (sense, docflow, chat, demo…)
├── runtime/              # Instance data (config, inbox, sessions — gitignored)
├── tests/                # Cross-cutting tests
├── docs/                 # Architecture specs, PRDs, design docs
├── infra/                # Terraform, Helm, K3D configs
├── scripts/              # Bootstrap and installation scripts
├── pyproject.toml        # Package definition (Hatchling)
├── Makefile              # Dev commands (test, build, lint, clean)
└── CLAUDE.md             # AI assistant context and project conventions

CLI

neut chat                              # Interactive agent
neut signal pipeline ingest             # Run signal ingestion
neut signal status                      # Pipeline health
neut pub publish                       # Generate and publish documents
neut doctor                            # AI-powered diagnostics
neut ext init my-extension             # Create a new extension
neut demo run collaborator             # Guided onboarding walkthrough

Development

make test              # Unit tests
make integration       # Integration tests (needs .env credentials)
make lint              # Ruff linter
make build             # Build wheel
make clean             # Remove build artifacts

Tech Stack

Layer	Technology
CLI & Agents	Python (argparse, rich, prompt-toolkit)
LLM Gateway	Model-agnostic (Anthropic, OpenAI, any OpenAI-compatible)
Data Platform	Apache Iceberg + DuckDB + Dagster + dbt
Database	PostgreSQL + pgvector
Infrastructure	Terraform, Helm, K3D
CI/CD	GitLab CI with AI code review
Packaging	Hatchling (pip-installable wheel)

Design Principles

• Reactor-agnostic core with facility-specific config
• Offline-first — queue locally, sync on restore
• Human-in-the-loop for all writes in safety-adjacent contexts
• No vendor lock-in — model-agnostic, cloud-agnostic, IDE-agnostic
• Provider pattern — swap implementations via config, not code changes

Contributing

See CONTRIBUTING.md for development workflow, git practices, and conventions.

License

MIT