dft-qe

Name: dft-qe
Author: jinzhezenggroup

113stars23forks

How to install dft-qe?

jinzhezenggroup/computational-chemistry-agent-skills/dft-qe

$npx -y skills add jinzhezenggroup/computational-chemistry-agent-skills --skill dft-qe

Installs into the current project.

Summary

Generate Quantum ESPRESSO DFT input tasks from a user-provided structure plus user-specified DFT settings. Use when the user wants to prepare QE calculations such as SCF, NSCF, relax, vc-relax, MD, bands, DOS, or phonons starting from a structure file or coordinates together with

Files · 1

View on GitHub

SKILL.md

1	# DFT with Quantum ESPRESSO
2
3	Use this skill to build a QE DFT task from a user-provided structure and DFT settings.
4
5	## Scope
6
7	This skill should:
8
9	- require a user-provided structure
10	- read or normalize the structure input
11	- identify the target calculation type
12	- collect the minimum required DFT settings
13	- generate the QE input file
14	- organize the task directory in a way that can be handed off to a submission skill
15	- state assumptions and unresolved choices
16
17	This skill should not:
18
19	- submit jobs
20	- manage schedulers
21	- handle remote execution
22	- invent critical scientific parameters
23
24	If the user wants the task submitted, hand off to another skill such as `dpdisp-submit` after the QE task is generated.
25
26	## Hard requirement
27
28	The user must provide a structure.
29
30	Do not generate a QE task without a user-provided structure source. If the structure is missing, stop and ask for it.
31
32	## Structure input convention
33
34	Use this bundled example-plus-layout pattern as the reference:
35
36	- generated QE input example: `assets/pw-water-0.in`
37	- user-provided structure staging pattern: `openclaw_input/` containing files such as `structure.xyz` and `CELL`
38
39	Treat this as the canonical pattern for what the user is expected to provide and what this skill is expected to generate.
40
41	The structure format does not need to be fixed to xyz. If the user provides another reasonable atomistic structure format, convert or normalize it as needed. When format conversion is needed, use the `dpdata-cli` skill.
42
43	For periodic systems, ensure cell information is preserved. If a plain xyz file is used, cell data must be provided separately, for example through a `CELL` file.
44
45	For a concrete file-oriented workflow, see `references/commands-and-workflow.md`.
46
47	## Expected workflow
48
49	1. Start from a user-provided structure.
50	1. Normalize the structure into the task layout if needed.
51	1. Determine the target QE calculation type.
52	1. Collect only the missing critical DFT parameters.
53	1. Generate the QE input file.
54	1. Place the generated task in a runnable task directory.
55	1. If submission is requested, pass that directory to `dpdisp-submit`.
56
57	## DFT parameters to collect
58
59	### Must provide
60
61	Do not generate a formal QE task unless these are known or explicitly confirmed:
62
63	- `calculation`
64	- `input_dft`
65	- `ecutwfc`
66	- `pseudo_dir`
67	- pseudopotential file for each element
68
69	### Usually should be explicit
70
71	These should normally be confirmed rather than guessed:
72
73	- `vdw_corr` when dispersion may matter
74	- `ecutrho` when relevant to the pseudopotential family or workflow
75	- `K_POINTS` setting, for example `gamma` or an automatic mesh
76	- occupation / smearing settings for metallic or ambiguous systems
77	- `conv_thr`
78	- `electron_maxstep`
79
80	### Task-specific additions
81
82	Also collect task-dependent parameters when relevant.
83
84	For `relax` / `vc-relax`:
85
86	- `forc_conv_thr`
87	- `etot_conv_thr`
88	- `cell_dofree` for `vc-relax`
89
90	For `md`:
91
92	- `nstep`
93	- `dt`
94	- `ion_temperature`
95	- `tempw`
96	- `ion_dynamics`
97
98	For spin-polarized or magnetic systems:
99
100	- `nspin`
101	- `starting_magnetization`
102	- charge or spin-related settings if requested
103
104	For advanced workflows if explicitly requested:
105
106	- Hubbard U settings
107	- hybrid-functional settings
108	- electric field settings such as `edir`, `emaxpos`, or related controls
109
110	Do not ask for everything at once; ask only for the missing essentials.
111
112	## Required behavior
113
114	1. Inspect the provided structure if accessible.
115	1. Determine elements, cell information, and coordinate representation.
116	1. If format normalization is needed, convert the structure using `dpdata-cli`.
117	1. Confirm the QE task type.
118	1. Gather only the missing critical DFT settings.
119	1. Generate the QE input yourself.
120	1. Explain assumptions clearly.
121	1. Flag unresolved scientific choices instead of hiding them.
122	1. Prepare the task directory so another skill can submit it.
123
124	## Template pattern
125
126	Use the QE example input included in this repository:
127
128	`assets/pw-water-0.in`
129
130	as the reference pattern for how a QE task is organized and how the `pw.x` input is laid out.
131
132	In this workflow:
133
134	- the user provides the structure in `openclaw_input`-style form
135	- this skill generate

People who liked this also viewed

6 picks

Type

dft-qe

113stars23forks

View on GitHub

How to install dft-qe?

jinzhezenggroup/computational-chemistry-agent-skills/dft-qe

$npx -y skills add jinzhezenggroup/computational-chemistry-agent-skills --skill dft-qe

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# DFT with Quantum ESPRESSO
2
3	Use this skill to build a QE DFT task from a user-provided structure and DFT settings.
4
5	## Scope
6
7	This skill should:
8
9	- require a user-provided structure
10	- read or normalize the structure input
11	- identify the target calculation type
12	- collect the minimum required DFT settings
13	- generate the QE input file
14	- organize the task directory in a way that can be handed off to a submission skill
15	- state assumptions and unresolved choices
16
17	This skill should not:
18
19	- submit jobs
20	- manage schedulers
21	- handle remote execution
22	- invent critical scientific parameters
23
24	If the user wants the task submitted, hand off to another skill such as `dpdisp-submit` after the QE task is generated.
25
26	## Hard requirement
27
28	The user must provide a structure.
29
30	Do not generate a QE task without a user-provided structure source. If the structure is missing, stop and ask for it.
31
32	## Structure input convention
33
34	Use this bundled example-plus-layout pattern as the reference:
35
36	- generated QE input example: `assets/pw-water-0.in`
37	- user-provided structure staging pattern: `openclaw_input/` containing files such as `structure.xyz` and `CELL`
38
39	Treat this as the canonical pattern for what the user is expected to provide and what this skill is expected to generate.
40
41	The structure format does not need to be fixed to xyz. If the user provides another reasonable atomistic structure format, convert or normalize it as needed. When format conversion is needed, use the `dpdata-cli` skill.
42
43	For periodic systems, ensure cell information is preserved. If a plain xyz file is used, cell data must be provided separately, for example through a `CELL` file.
44
45	For a concrete file-oriented workflow, see `references/commands-and-workflow.md`.
46
47	## Expected workflow
48
49	1. Start from a user-provided structure.
50	1. Normalize the structure into the task layout if needed.
51	1. Determine the target QE calculation type.
52	1. Collect only the missing critical DFT parameters.
53	1. Generate the QE input file.
54	1. Place the generated task in a runnable task directory.
55	1. If submission is requested, pass that directory to `dpdisp-submit`.
56
57	## DFT parameters to collect
58
59	### Must provide
60
61	Do not generate a formal QE task unless these are known or explicitly confirmed:
62
63	- `calculation`
64	- `input_dft`
65	- `ecutwfc`
66	- `pseudo_dir`
67	- pseudopotential file for each element
68
69	### Usually should be explicit
70
71	These should normally be confirmed rather than guessed:
72
73	- `vdw_corr` when dispersion may matter
74	- `ecutrho` when relevant to the pseudopotential family or workflow
75	- `K_POINTS` setting, for example `gamma` or an automatic mesh
76	- occupation / smearing settings for metallic or ambiguous systems
77	- `conv_thr`
78	- `electron_maxstep`
79
80	### Task-specific additions
81
82	Also collect task-dependent parameters when relevant.
83
84	For `relax` / `vc-relax`:
85
86	- `forc_conv_thr`
87	- `etot_conv_thr`
88	- `cell_dofree` for `vc-relax`
89
90	For `md`:
91
92	- `nstep`
93	- `dt`
94	- `ion_temperature`
95	- `tempw`
96	- `ion_dynamics`
97
98	For spin-polarized or magnetic systems:
99
100	- `nspin`
101	- `starting_magnetization`
102	- charge or spin-related settings if requested
103
104	For advanced workflows if explicitly requested:
105
106	- Hubbard U settings
107	- hybrid-functional settings
108	- electric field settings such as `edir`, `emaxpos`, or related controls
109
110	Do not ask for everything at once; ask only for the missing essentials.
111
112	## Required behavior
113
114	1. Inspect the provided structure if accessible.
115	1. Determine elements, cell information, and coordinate representation.
116	1. If format normalization is needed, convert the structure using `dpdata-cli`.
117	1. Confirm the QE task type.
118	1. Gather only the missing critical DFT settings.
119	1. Generate the QE input yourself.
120	1. Explain assumptions clearly.
121	1. Flag unresolved scientific choices instead of hiding them.
122	1. Prepare the task directory so another skill can submit it.
123
124	## Template pattern
125
126	Use the QE example input included in this repository:
127
128	`assets/pw-water-0.in`
129
130	as the reference pattern for how a QE task is organized and how the `pw.x` input is laid out.
131
132	In this workflow:
133
134	- the user provides the structure in `openclaw_input`-style form
135	- this skill generate

People who liked this also viewed

6 picks

Type