菜单
Find and acquire computational science resources autonomously. Use when you need force field parameters, pseudopotentials, crystal structures, or any other scientific data. You are a researcher - you find what you need.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | resource-acquisition |
| description | Find and acquire computational science resources autonomously. Use when you need force field parameters, pseudopotentials, crystal structures, or any other scientific data. You are a researcher - you find what you need. |
| allowed-tools | ["Read","Write","Bash","Glob","Grep","WebSearch","WebFetch"] |
You are a researcher. You have tools, you have a goal, you figure out the rest.
When you need something (parameters, structures, files):
Never use a value without knowing where it came from.
Every MD simulation needs force field parameters (LJ epsilon/sigma, bond constants, etc.). These are NOT universal - they depend on:
Step 1: Identify what you need
"I need Lennard-Jones parameters for liquid argon at 94.4 K"
"I need TIP4P water model parameters"
"I need EAM potential for copper"
Step 2: Search literature
Search queries that work:
- "[material] lennard-jones parameters molecular dynamics"
- "[material] force field parameters"
- "[model name] original paper" (e.g., "TIP4P original paper")
- "[material] interatomic potential"
Step 3: Find the authoritative source For common systems, there are seminal papers:
Step 4: Extract parameters
Step 5: Convert units if necessary Common conversions:
Step 6: Document your source In your input file:
# Lennard-Jones parameters for argon
# Source: Rahman, Phys. Rev. 136, A405 (1964)
# ε = 0.238 kcal/mol, σ = 3.405 Å
pair_coeff 1 1 0.238 3.405
1. Search: "argon lennard-jones parameters molecular dynamics"
2. Find: Rahman 1964 is the seminal paper for liquid Ar MD
3. Also find: Allen & Tildesley give ε/kB = 119.8 K, σ = 3.405 Å
4. Convert: ε = 119.8 K × 0.001987 kcal/mol/K = 0.238 kcal/mol
5. Use: pair_coeff 1 1 0.238 3.405
QE needs pseudopotential files (.UPF) for each element. These depend on:
Primary Sources (in order of preference):
SSSP (Standard Solid State Pseudopotentials)
PseudoDojo
QE Pseudopotential Library
Materials Cloud
Step 1: Determine what you need
Element: Si
Functional: PBE (most common for solids)
Type: Usually US or PAW for efficiency
Step 2: Search and navigate
Use WebSearch: "silicon PBE pseudopotential SSSP"
Or navigate directly to SSSP table
Step 3: Download the file Use Playwright or WebFetch to download:
The file will be something like:
Si.pbe-n-rrkjus_psl.1.0.0.UPF
Step 4: Save to resources directory
Save to: workspaces/resources/pseudopotentials/
Or to your project workspace
Step 5: Reference in input
ATOMIC_SPECIES
Si 28.0855 Si.pbe-n-rrkjus_psl.1.0.0.UPF
When you download a pseudopotential, also note the recommended cutoffs:
SSSP provides these explicitly. If not available, test convergence.
Materials Project (API available)
Crystallography Open Database (COD)
ICSD (subscription required)
Paper Supplementary Information
From Materials Project:
from mp_api.client import MPRester
import os
api_key = os.environ.get("MP_API_KEY")
with MPRester(api_key) as mpr:
structure = mpr.get_structure_by_material_id("mp-149") # Silicon
structure.to("poscar", "POSCAR") # Save as VASP format
From COD or papers:
from pymatgen.core import Structure
struct = Structure.from_file("structure.cif")
The main paper often says "parameters in SI" or "see Supporting Information". You need to get these files.
Step 1: Find the paper DOI From Semantic Scholar, Google Scholar, or the paper itself.
Step 2: Navigate to publisher page Use Playwright to:
Step 3: Parse the SI
1. Search Semantic Scholar for "TIP4P water Jorgensen 1983"
2. Get DOI: 10.1063/1.445869
3. Navigate to: https://doi.org/10.1063/1.445869
4. Find paper, check if SI exists
5. For this classic paper, parameters are in Table I of main text
6. Extract: ε = 0.1550 kcal/mol, σ = 3.1536 Å, etc.
When you find parameters:
Check that parameters make sense:
workspaces/resources/
├── pseudopotentials/
│ ├── pbe/
│ │ ├── Si.pbe-n-rrkjus_psl.1.0.0.UPF
│ │ └── ...
│ └── lda/
├── potentials/
│ ├── eam/
│ └── tersoff/
├── structures/
│ ├── cif/
│ └── poscar/
└── parameters/
└── force_fields.json # Cache of found parameters
When you find parameters, save them:
{
"argon_lj": {
"epsilon_kcal_mol": 0.238,
"sigma_angstrom": 3.405,
"source": "Rahman 1964, Phys. Rev. 136, A405",
"notes": "For liquid argon near triple point"
}
}
You are a researcher, not a script executor.
The goal is: given only a scientific question, you acquire everything needed to answer it.
Get fully up to date on a project, or build its weekly / PI-meeting update bundle, from INSIDE the project directory. Use when an agent lands cold in a project and needs the current state (open workstreams, what moved, open PI decisions, blockers), or when asked to "build the weekly update", "prep the PI meeting", or "aggregate the relevant files into the update dir". Mines only THIS repo, driven by its own .sync/manifest.yaml. No dependency on LabSync.
Run Quantum ESPRESSO DFT calculations. Use when asked to perform first-principles calculations, SCF, structural relaxation, band structure, DOS, phonons, or any ab initio quantum mechanical calculation.
Decide where and how to run compute jobs across available backends (HPC, cloud GPUs, local). Use when starting any compute-intensive task, choosing a backend or partition, deciding job parameters (walltime, GPU type, parallelism), or debugging compute infrastructure issues. Embeds the smoke-first iteration pattern — validate cheaply before scaling.
Orchestrate one or more long-running compute campaigns (HPC simulations, multi-stage MD pipelines, ensemble runs) where stages take hours to days and failures must not halt the pipeline. Use when running multiple parallel campaigns that share a stage structure (smoke → equilibration → production → analysis), when waiting on long jobs makes interactive driving impractical, or when designing a self-healing tick-based supervisor for compute work. Reads/writes per-campaign WORKFLOW.md files as durable state.
Analyze simulation data and compute properties. Use when asked to parse LAMMPS/QE output, calculate diffusion coefficients, RDF, MSD, energies, or generate plots and visualizations.
Generate candidate crystal structures from a chemical formula and assess their stability using ggen. Use when you need trial structures for a composition with no known experimental structure, or want stability-ranked polymorph candidates before running MD/DFT.