Codex skill for turning local experimental data folders into clean datasets, publication-style figures, and editable Origin .opju projects.
It is designed for materials and chemistry workflows where a data folder may contain XRD, FTIR, SEM-EDS reports, TG/TGA, LOI, cone calorimetry, contact angle, antifungal images, and mechanical tests.
- Recursively scans a local data folder.
- Identifies useful numeric data and reports.
- Exports cleaned CSV/XLSX data tables.
- Creates editable Origin workbooks and graphs through Origin COM.
- Adds valid error bars when replicate or report error data exist.
- Avoids fake error bars for single-run spectra such as XRD and FTIR.
Default Origin graph style:
- Bottom/left tick labels: 36 pt, bold.
- Bottom/left axis titles: 40 pt, bold.
- Full box border.
- Top/right axes keep border only, with no ticks, labels, or titles.
- Border width: 3.
Copy the skill folder into your Codex skills directory:
Copy-Item -Recurse .\data-origin-plotter "$env:USERPROFILE\.codex\skills\data-origin-plotter"Then restart Codex or refresh skills.
Ask Codex something like:
Use data-origin-plotter. My data folder is D:\experiment\TiO2-2. Draw the usable figures and export an Origin file.
- Windows.
- Origin/OriginPro installed and COM automation available.
- For Origin export,
Origin.ApplicationSIshould start successfully.
If Origin COM hangs, open Origin manually once, close welcome/license/update dialogs, save and close a blank project, then retry.
data-origin-plotter/
├── SKILL.md
├── agents/
│ └── openai.yaml
└── scripts/
└── build_origin_project.ps1
This skill intentionally does not include experimental data. Keep private research folders and generated Origin files out of the repository unless you explicitly want to publish them.