Access — SCPN-Control

Install from PyPI

# Core (numpy, scipy, click)
pip install scpn-control

# With visualisation and dashboard
pip install "scpn-control[viz,dashboard]"

# With JAX autodiff + machine learning
pip install "scpn-control[jax,ml]"

# With reinforcement learning
pip install "scpn-control[rl]"

# Everything
pip install "scpn-control[all]"

Requires Python ≥3.10. Rust extension (PyO3) ships as pre-built wheels for Linux x86_64 and macOS ARM64. Source build requires Rust ≥1.75 and maturin.

Public Repository

Source and evidence are public

Inspect, clone, and review the full source at github.com/anulum/scpn-control. The repository carries the validation scripts, evidence gates, documentation, and contribution workflow alongside the package.

Use and contribution routes

CommercialLicence

Commercial licensing and support are available for proprietary integration; source visibility itself does not require purchase. See pricing.

AcademicFree

University groups and publicly funded laboratories can use the public AGPL source directly and cite the exact release or commit used.

ContributorsCLA

Open-source contributors can fork the public repository and open a pull request under the documented contribution and CLA process.

ReviewersNDA

Reviewers can inspect the public source and evidence directly; contact ANULUM only if a review requires non-public supporting material.

Citation

Academic BibTeX

@software{scpn_control,
  author = {Šotek, Miroslav},
  title = {SCPN-Control: Neuro-Symbolic Stochastic Petri Net
          Control Engine for Tokamak Plasma},
  year = {2026},
  publisher = {ANULUM},
  url = {https://pypi.org/project/scpn-control/}
}

JOSS Paper

Journal of Open Source Software

A JOSS paper describing the architecture, physics implementations, and validation methodology is in preparation. Pre-print will be available on arXiv prior to submission. DOI will be added here upon acceptance.

Contributing

Contributor Licence Agreement

All contributions require a signed CLA granting ANULUM the right to relicence contributions under both AGPL-3.0 and commercial terms. This is standard dual-licensing practice — your contributions remain attributed to you, and you retain copyright.

Contribution areas of interest: plasma physics models, Rust acceleration, controller algorithms, test coverage, documentation. We do not accept AI-generated code without human review and attestation.

Scope Notice

SCPN-Control is alpha-stage research software developed by a single researcher. It implements publication-accurate physics but has not been validated on real tokamak hardware. It is not a production Plasma Control System. Do not use it for safety-critical applications without independent verification, institutional review, and regulatory approval.