raman-spectra-simulation

name	raman-spectra-simulation
description	Compute and visualize Raman Spectra from input SMILES or XYZ file with MLatom.

Raman Spectra Simulation Skill

When to use this

Use this skill when the user provides a SMILES or XYZ file (coordinates) and wants:

• Optimized geometry
• Theoretically calculated Raman spectrum in a summarized table
• Quick overview of simulated Raman spectrum
• Publication-ready image file

Inputs

• SMILES string (e.g. CCO for ethanol) or XYZ file path (necessary)
• A corresponding experimental Raman spectrum (optional, .txt format)

Outputs

• /tmp/chemclaw/optimized_geometry.md
• /tmp/chemclaw/theoretical_raman.md
• /tmp/chemclaw/theoretical_raman.png
• /tmp/chemclaw/exp_raman.png (optional, if experimental data provided)
• /tmp/chemclaw/comparison_raman.png (optional, if experimental data provided)

Agent response

When returning results to the user, include the generated spectrum image(s) in the response so the user sees them directly. Use markdown image syntax with the correct path to the PNG file(s), e.g. ![Raman spectrum](path/to/theoretical_raman.png). Include comparison_raman.png if experimental data was provided.

New environment (from zero)

conda create -n spec python=3.12 -y
conda activate spec
conda install -c conda-forge xtb -y   # macOS 必装（MLatom 内建 xTB 为 Linux 版）
pip install mlatom numpy matplotlib pyscf geometric rdkit

cd raman-spectra-simulation
python raman_spectra_simulation.py CCO
# optional: python raman_spectra_simulation.py /path/to/structure.xyz
# optional: python raman_spectra_simulation.py CCO experimental_raman.txt

• 请在 skill 目录内执行脚本。
• macOS 若仍误用 MLatom 内建 Linux 版 xTB，可 export PATH="$CONDA_PREFIX/bin:$PATH" 让系统先找到 conda 的 xtb。

How to use（已装好环境时）

SMILES:

cd raman-spectra-simulation
python raman_spectra_simulation.py CCO

XYZ（用户提供坐标文件）:

cd raman-spectra-simulation
python raman_spectra_simulation.py /path/to/structure.xyz

含实验谱对照:

cd raman-spectra-simulation
python raman_spectra_simulation.py CCO experimental_raman.txt