Run simulations
Launch packaged scenarios, demos, and Campaign Studio studies from the CLI or local cockpit.
Public beta | Apache-2.0 | PyPI + Zenodo archived
Run simulations. Inspect evidence. Compare campaigns. Publish reproducible artifacts.
QS-DMSS is a deterministic, evidence-first simulation lab for QuantumScalar dark matter workflows. It turns a run into a research object: configured, measured, bundled, verified, replayable, comparable, citable, and ready for review.
The QS-DMSS loop
A simulation is not finished when it produces a plot.
Inspect evidence
Open metrics, manifests, reports, CSV samples, figures, verification status, and replay paths.
Compare campaigns
Evaluate parameter grids against explicit objectives, constraints, ranking weights, and recommendations.
Publish artifacts
Export research objects with citation blocks, evidence bundles, scenario metadata, and reproducibility notes.
Studio capabilities
Built around evidence, not just execution.
Lab Mode
A guided cockpit path for choosing a scenario, running the simulation, reading plain-language interpretation, and inspecting generated evidence.
Campaign Studio
Reusable study templates, editable parameter grids, decision profiles, scoring contracts, recommendation rationale, and last-run provenance.
Evidence Bundles
Run directories preserve configs, metrics, environment locks, figures, manifests, reports, hashes, verification status, and replay commands.
Publication Export
Compose a shareable research-object surface with scenario context, artifacts, metrics, evidence status, replay guidance, and DOI citation.
Dry-Run Slurm Review
Generate scheduler request bundles for HPC/RSE review without calling sbatch, squeue, or any real cluster submission path.
Scientific Boundaries
QS-DMSS is beta software for reproducible package/evidence workflows. It does not claim peer-reviewed physical validation or production cosmological simulation performance.
Try it locally
Run the cockpit from the published package.
The current safest interactive experience is local-first. Install from PyPI, start the cockpit on localhost, then run Lab Mode or Campaign Studio without exposing a public compute surface.
python -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
python -m pip install qs-dmss==0.9.0
qs-dmss cockpit --host 127.0.0.1 --port 8001Phase two
Live demo coming next at app.qs-dmss.studio.
The future hosted app should be a constrained demonstration, not an unrestricted public compute service. The first version should allow only packaged scenarios and study templates, short-lived artifacts, no arbitrary YAML upload, no user-provided filesystem paths, no real scheduler submission, and clear rate/resource limits.