| name | trajectory-analysis |
| description | Use this skill for generic CP2K/LAMMPS trajectory and run-health checks and for deciding when task-specific trajectory parsing is required. |
| allowed-tools | md_trajectory_summary analyze_trajectory lammps_log_summary cp2k_output_summary execute |
trajectory-analysis
Overview
Use this skill for generic MD health summaries. It is not a replacement for task-specific mechanistic analysis.
Quick Start
- Locate the result directory or trajectory file.
- Use
md_trajectory_summary for frame counts, final frame export, restart-file presence, generic thermo/log evidence, CP2K .ener files, and LAMMPS RDF/MSD tables.
- Use
analyze_trajectory for ASE-readable trajectories when MSD/RDF artifacts are requested.
- Write a focused parser under
scripts/ for system-specific residence, reaction, adsorption, or free-energy questions.
Allowed tools
md_trajectory_summary
analyze_trajectory
lammps_log_summary
cp2k_output_summary
execute
Workflow
1. Start with health, not interpretation
- Check frame count, atom count, final frame export, time span if recoverable, log completion, energy/thermo files, and restart files.
- Inspect temperature, energy, and pressure drift before interpreting a trajectory.
2. Keep observables explicit
- RDF and MSD require explicit species or group choices; if they were produced by in-run LAMMPS computes, summarize file presence and numeric rows before interpretation.
- Mechanistic labels such as diffusion path, desorption event, residence time, or reaction coordinate usually require a task-specific script.
Method-critical defaults
- Generic summaries may prove that a run produced data; they do not prove equilibration or scientific convergence.
- Do not infer barriers or free energies from raw MD without the appropriate method and parser.
Output Contract
Return:
- trajectory file path
- summary JSON path
- log or CP2K output summary path if available
- any task-specific parser path
- explicit limitations
References