openmm-system-minimization— agent skill

Purpose

Provide a minimal, runnable OpenMM starter that creates a small bonded system, minimizes its energy, and reports before/after energies and coordinates.

When to use

• You need a verified local OpenMM entry point.
• You want a tiny deterministic minimization example before scaling to larger simulations.

When not to use

• You need biomolecular force-field preparation or explicit-solvent production runs.
• You need ensemble sampling or long-timescale trajectories.

Inputs

• Optional JSON summary path

Outputs

• JSON summary with OpenMM version, platform, initial/final potential energy, and final coordinates

Requirements

• slurm/envs/chem-tools with OpenMM available

Procedure

1. Run slurm/envs/chem-tools/bin/python skills/computational-chemistry-and-molecular-simulation/openmm-system-minimization/scripts/run_openmm_minimization.py.
2. Compare initial_potential_energy_kj_mol and final_potential_energy_kj_mol.
3. Reuse the script as a seed for more realistic system-building workflows.

Validation

• The command exits successfully.
• Final potential energy is lower than the initial potential energy.
• Final coordinates are emitted for all particles.

Failure modes and fixes

• Missing OpenMM environment: run the script with slurm/envs/chem-tools/bin/python.
• Numerical issues: reset the toy system to the bundled defaults.

Safety and limits

• Educational starter only.
• This is not a physically meaningful production system.

Provenance

• OpenMM user guide: https://docs.openmm.org/latest/userguide/

Related skills

• rdkit-molecular-descriptors