Collaboration request: plasma-control and facility replay artefacts

[anulum]

Blocked modules

• src/scpn_control/control/gym_tokamak_env.py
• src/scpn_control/control/rzip_model.py
• src/scpn_control/control/rwm_feedback.py
• src/scpn_control/control/free_boundary_tracking.py
• src/scpn_control/control/density_controller.py
• src/scpn_control/control/tokamak_digital_twin.py
• src/scpn_control/control/advanced_soc_fusion_learning.py
• src/scpn_control/control/burn_controller.py
• src/scpn_control/control/volt_second_manager.py
• src/scpn_control/control/mu_synthesis.py
• src/scpn_control/control

Needed evidence

• Higher-fidelity replay cases and plant-model validity bounds for reduced-order training environments.
• Facility parameter sources, reference RZIP/CREATE-L/NL/TSC benchmarks, or measured vertical-displacement benchmarks.
• Measured RWM shots or external MHD stability references with wall geometry, rotation, coil coupling, sensor geometry, latency, and beta-limit provenance.
• Measured or validated free-boundary replay cases with actuator, latency, and sensor-calibration evidence.
• Equilibrium-derived flux-surface volumes, recycling-source evidence, density profile evolution, and fuelling cases.
• Replayed or measured discharge histories for digital twin IDS export and actuator-lag summaries.
• Gyrokinetic or measured turbulence evidence and replayed plasma-control objectives for SOC learning policies.
• Integrated transport or published burn-control benchmarks and calibrated alpha-slowing/ash-removal evidence.
• Real loop-voltage/current traces and neoclassical or transport-solver bootstrap evidence for volt-second claims.
• Validated frequency-dependent H-infinity backend, benchmark plants, frequency grids, D-scale fits, and mu bounds.

Claim unblocked

Facility-control, deployment-readiness, and stronger reduced-order plant claims. Until then, these surfaces remain bounded-control models.