L’interface est localisée ; les présentations et documents des Skills sont traduits automatiquement par IA dans votre langue. Le code, les commandes, chemins et termes techniques restent en version originale, et le document de l’auteur fait foi.
Explorer les Skills
Productivité · SKILL.md

smirnoff-force-fields

smirnoff-force-fields

Load, modify, inspect, serialize, label, and apply SMIRNOFF OFFXML force fields with OpenFF Toolkit ForceField APIs.

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentation de la Skill

GitHub

name: smirnoff-force-fields description: "Load, modify, inspect, serialize, label, and apply SMIRNOFF OFFXML force fields with OpenFF Toolkit ForceField APIs." disable-model-invocation: true metadata: disco-role: operating license: MIT

SMIRNOFF Force Fields

Use this sub-skill when a task centers on OpenFF Toolkit SMIRNOFF force fields: loading .offxml files, discovering installed force fields, inspecting or modifying parameter handlers, labeling molecules with assigned parameters, serializing edited force fields, combining sources, creating Interchange/OpenMM systems, or diagnosing SMIRNOFF parsing and handler errors.

Route First

  • Use references/api-reference.md for the verified ForceField signatures, handler concepts, label shapes, serialization APIs, and Interchange/OpenMM boundary.
  • Use references/workflows.md for copy-ready recipes: load installed or local .offxml, inspect available force fields, label molecules, edit a parameter, serialize/reload, and apply a force field.
  • Use references/plugins-and-custom-parameters.md for custom ParameterHandler classes and entry-point plugin loading with load_plugins=True.
  • Use references/troubleshooting.md when parsing, handler registration, force-field package discovery, charge assignment, virtual-site, Interchange, or OpenMM conversion fails.
  • Use scripts/smoke_smirnoff_workflow.py for a safe local smoke probe that builds a molecule/topology, loads a force field, optionally labels parameters, and attempts Interchange if installed.

Boundaries

  • For molecule construction, stereochemistry, conformers, file I/O, and mapped SMILES details, route to ../molecules-and-io/SKILL.md.
  • For topology construction, PDB/OpenMM topology conversion, positions, boxes, and system-level preparation depth, route to ../topology-and-systems/SKILL.md.
  • For optional backend installation, toolkit registries, charge toolkit availability, and dependency selection depth, route to ../toolkit-backends/SKILL.md.
  • Do not rely on original repository docs, tests, examples, or notebooks at runtime; this sub-skill bundles the SMIRNOFF patterns needed for future agents.

Quick Start

from openff.toolkit import ForceField, Molecule, Topology
from openff.toolkit.typing.engines.smirnoff import get_available_force_fields

print(get_available_force_fields())
force_field = ForceField("openff-2.3.0.offxml")
molecule = Molecule.from_smiles("CCO")
topology = Topology.from_molecules([molecule])
labels = force_field.label_molecules(topology)
print(force_field.registered_parameter_handlers)
print(labels[0].keys())

If ForceField("openff-2.3.0.offxml") cannot find the file, inspect get_available_force_fields() and install or point to a local .offxml file rather than changing the molecule/topology workflow.

Skills enregistrées