Upload profiling figures

[max-models]

This PR uploads all figures found in results_dir = os.path.join(sim.env.path_out, "results") to the profiling-data repository. These figures are then displayed on the individual run pages.

Example: https://struphy-hub.github.io/profiling-data/run/?case=20260803T090820Z-b8a5ed8f-poisson_cube_strong_scaling&run=run03#gallery

Additionally, I cleaned up the filenames and folder structure in profiling-data, now all the files for each run are categorized in a case/results-run02. See here.

[Copilot]

Pull request overview

This PR extends the profiling packaging/upload flow so that each run’s post-processed artifacts (figures/arrays written to <sim_dir>/results) are included in the packaged case folder and uploaded to the struphy-hub/profiling-data repository, with a cleaned-up per-run folder structure (<case>/results-runXX).

Changes:

• Add standalone CLI support to upload an already-packaged case directory to profiling-data.
• Restructure packaged output to be per-run (results-runXX/) and write per-run metadata files (runXX.json) referenced from case_metadata.json.
• Update the Poisson profiling example to write post-processed artifacts to sim.env.path_out/results so packaging can pick them up.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 3 comments.

File	Description
profiling/upload.py	Adds a CLI entrypoint to upload an already-packaged case folder to profiling-data using existing clone/push helpers.
profiling/profiling_job.py	Switches packaging bookkeeping from per-file to per-run, copies sim_<id>/results artifacts, and updates finalize messaging.
profiling/package_profiling_results.py	Introduces per-run naming helpers, per-run metadata writer, and logic to copy result artifacts from <sim_dir>/results.
profiling/examples/Poisson/cube_strong_scaling/params_poisson.py	Writes post-processed outputs into sim.env.path_out/results to align with the new packaging collector.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.