Vasp analysis
Agent skill for VASP result analysis, including energies, DOS, bands, magnetism, voltage, and NEB.
npx -y skills add chatmaterials/vasp-analysisAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use when the task is to analyze completed or partially completed VASP results, including energy extraction, convergence interpretation, structure or force summaries, comparing multiple VASP runs, and writing compact reports from OUTCAR, OSZICAR, INCAR, or POSCAR.
SKILL.md
1.7 KB, as published. Nobody here has run it
VASP Analysis
Use this skill for post-run VASP result analysis rather than workflow setup.
When to use
- analyze a completed or incomplete VASP run
- compare energies across multiple VASP directories
- summarize convergence, forces, or status from
OUTCARandOSZICAR - write a compact status or comparison report from existing results
Use the bundled helpers
scripts/analyze_vasp_result.pySummarize a single VASP result directory.scripts/compare_vasp_results.pyCompare multiple VASP result directories by energy and status.scripts/structure_change_vasp.pyCompare two VASP structures and summarize lattice or coordinate changes.scripts/analyze_vasp_dos.pyExtract DOS-oriented summary data fromDOSCAR.scripts/analyze_vasp_band.pyExtract band-gap-oriented summary data fromEIGENVAL.scripts/analyze_vasp_magnetism.pyExtract total and local magnetic moments fromOUTCAR.scripts/analyze_vasp_voltage.pyEstimate an average insertion voltage from two VASP states.scripts/analyze_vasp_neb.pyEstimate a simple NEB barrier from numbered VASP image directories.scripts/export_analysis_report.pyExport a markdown analysis report from a VASP result directory.
Guardrails
- Do not claim a scientific conclusion from an incomplete or unconverged run.
- Distinguish raw extraction from interpretation.
- If directories are not numerically comparable, say so plainly.