DFT Research Workflow is an English, software-neutral, human-first research manual for carrying out a density-functional-theory study.
The site starts with obtaining or building a material structure, continues through calculation preparation and reference-state calculations, branches into specific target calculations, and ends with analysis, validation, documentation, and preservation.
The reader-facing structure is:
A · Structure & Model
B · Method & Numerical Setup
C · Reference State
D · Target Calculations
E · Validation, Interpretation & Reproducibility
A, B, C, and E provide the common research backbone. D is a branching library in which band structures, density of states, phonons, defect energetics, electron–phonon coupling, superconductivity, transport, optical calculations, and other concrete calculations may each have an independent page.
workflow/topics.json defines stable A–E sections, D1–D5 groups, public topic
slugs and titles, and internal migration references. The home page, Research
Workflow directory, and topic routes are generated from this registry. The
registry length is not presented as the number of DFT operations.
Earlier O01–O24 and Operation 00–34 records remain temporarily as migration and URL-compatibility data, not as competing public taxonomies.
Every scientific page is organized naturally according to its subject. The site does not require a repeated Inputs/Outputs/Requirement-style contract or a uniform visible heading sequence.
The public navigation is Home, Research Workflow, Worked Workflows,
and Tools. Cross-cutting Framework material now lives in the A–E topics it
supports; the old /framework/* routes are migration surfaces. Likewise,
/recipes/* preserves old links while /workflows/ publishes only workflows
with a continuous, file-backed execution chain.
Public site: https://maxwell3919.github.io/DFT-Research-Workflow/
workflow/
├── topics.json # current A–E topic and route authority
└── practical-evidence.json # guide evidence class and case binding
examples/cases/
└── <case-id>/ # reproducible inputs, outputs and evidence gates
docs/
├── architecture.md
├── content-contract.md
└── content-population-plan.md
src/content/
├── topics/ # optional natural topic narratives
├── recipes/ # transitional workflow bindings
├── framework/ # sources for framework migration surfaces
├── core-operations/ # transitional O01–O24 bindings
└── operations/ # former 00–34 route metadata
ontology/
├── operations.json # transitional O01–O24 source
├── relations.json # transitional relation source
├── tags.json # contextual vocabulary
└── legacy-operations.json # former 00–34 route mapping
recipes/
└── index.json # transitional workflow source
See docs/architecture.md for the current A–E information architecture and migration policy. See docs/content-contract.md for natural writing, review, source, registry-binding, and migration rules. See docs/content-population-plan.md for the active writing and migration sequence.
Node.js 22.12 or newer is required.
npm ci --no-audit --no-fund
npm run dev
npm run checknpm run check validates the current topic registry, transitional migration
sources, project-base-safe links, content policy, case schemas and
hashes, Astro types, the static build, and generated routes. To re-run every
committed case extractor, parser, and gate checker (including verifying that a
declared FAIL returns nonzero), use:
CASE_PYTHON="$(pwd)/.venv/bin/python" node scripts/validate-cases.mjs --executeFor browser validation:
npm run preview -- --host 127.0.0.1 --port 4322
npm run smoke:browserRepository, build, and browser checks validate software structure and public behaviour. They do not validate any real DFT protocol, numerical result, physical interpretation, or scientific claim.