Matthias Volk

Check the (maximal) probability to reach a stable configuration. mCRL2 gives the expected result 1, equal to the [Prism model](https://www.prismmodelchecker.org/casestudies/self-stabilisation.php#ij).

- Aligned BRP model a bit closer with the [Prism model](https://www.prismmodelchecker.org/casestudies/brp.php) - Added two properties to check. mCRL2 gives the same results as Prism/Storm.

CI for weekly wheels

7

The original workflow `create_wheel.yml` needs the Storm version as input argument. This wrapper workflow is scheduled each week to build the wheels for the current master of Storm.

Upgrade to pybind version 3.0

4

See the pybind11 [upgrade guide](https://pybind11.readthedocs.io/en/stable/upgrade.html#v3-0) and the [Changelog](https://pybind11.readthedocs.io/en/stable/changelog.html#final-july-10-2025). Main steps as far as I can see are:

Ensure consistency between Jupyter notebooks and scripts in examples folder

1

Triggered by #255, I was wondering whether there is a good way to ensure that the scripts in the examples folder are actually consistent with the documentation in the Juypter...

Keeping track of documentation issues from pybind11. - Documentation for static or read-only members is not generated. This is an issue with pybind and needs fixing on their side. See...

The pure Python methods should have type information.

- Fetch tests should only be run on pushes to master and in weekly tests - Tests for installed Storm version should be run on PRs.

Should be able following [this documentation](https://docs.pytest.org/en/stable/explanation/goodpractices.html#tests-as-part-of-application-code)

The model files (Prism, Jani, etc) are currently part of `lib/stormpy/examples` and therefore also part of the wheel. We should move them to resources instead.