Hub on GitHub
AN OPEN SCIENTIFIC SOFTWARE INITIATIVE

One parameter layer.
Many scientific codes.

SNT Hub brings scientific codes and reusable DIPL parameter models together. Explore a code's inputs, prepare a local example, and see what its integration has actually been tested to do.

Explore the registry See how it works
Starting October 2026 Open source Optional integration Evidence led
✦
✧
model.dip
simulation
  box_size float = 2.5 m
    !condition ({.} > 0 m)
  steps int = 128
typed · checked · translated
↗native input
BoxSize   2.5
MaxSteps  128
...
Ready for the solver
DIPL → SNT → CODE
SEMANTIC INPUTS✳PHYSICAL UNITS✳VALIDATION✳NATIVE OUTPUTS✳REPRODUCIBILITY
HOW IT WORKS

A clearer path
to simulation inputs.

Each Hub project pairs a pinned scientific code with a DIPL model of selected inputs and a documented way to prepare them.

01 / MODEL
{ }

Describe the science

DIPL keeps values, units, types, relationships, and constraints together in a portable definition.

02 / CHECK
✓

Make inputs inspectable

SNT evaluates the model and catches declared errors before the settings reach a solver.

03 / PREPARE
↗

Generate native inputs

Use snt hub to fetch a pinned project and set up an example. Its adapter prepares files for the existing solver.

Why use SNT Hub? See where the extra layer pays off
CHOOSE A PROJECT

Explore integrations.

Browse 5 projects with example inputs and parameter references. Each listing separates how SNT connects, who supports the work, and what has been validated.

View the registry
THE INTEGRATION CAN GROW

Start with native files.
Go further when ready.

The adapter is an initial integration. A code's maintainers can later choose direct SNT3 access or generated static parameters while preserving a documented, inspectable model.

PATH 01 / DIRECT USE

Use evaluated parameters

Connect a code to SNT3 through its C++ or Python APIs, or evaluate the experimental C binding where appropriate.

C++ guide ↗Python guide ↗
PATH 02 / STATIC EXPORT

Generate native parameters

Export validated environments to C++, C, Fortran, Rust, Julia, JSON, or YAML for applications that should not parse DIPL at run time.

See static generation ↗
Inspect the model at any stage.

The optional, read-only Parameter Viewer shows evaluated values, units, provenance, and dependencies from a live DIPL project or DIPH5 snapshot.

Explore the viewer ↗
GO DEEPER

Explore the ecosystem.

Start with the code, read the initiative, or get into the DIPL language and SNT3 tools.

01 / SOURCESNT3 on GitHub

Explore the implementation, examples, and releases.

02 / INITIATIVEHow to participate

See the integration path and the meaning of registry statuses.

03 / LANGUAGEDIPL specification

Learn the syntax for typed, unit-aware scientific inputs.

04 / GUIDESSNT3 documentation

Follow tutorials and API guides for using the toolkit.

FOR PROJECT CONTRIBUTORSHub integration specification

Bundle requirements and recommended DIPL node, schema, and example organization.

Download PDF ↗Read Markdown ↗
STARTED10/26OCTOBER 2026
AN OPEN BEGINNING

Help grow
the next chapter.

SNT Hub began in October 2026. The registry is a starting collection, and its coverage and evidence will grow through new models, code-owner reviews, and reproducible checks. Researchers and maintainers can contribute a project or help assess an existing one.

Add a project Start a discussion