A web app for drafting CRediT (Contributor Roles Taxonomy) contribution statements for scholarly publications.
Add contributors, assign the 14 roles, and copy a manuscript-ready statement. It also builds a contribution heatmap and exports for journal submission systems: JATS4R XML, CSV, JSON, and Markdown.
Try it: credit.duinlab.nl
- Contributors: add, rename, reorder, paste a whole author list, or paste an ORCID iD or URL to look up the name
- Contribution grid: click cells to assign the 14 roles as yes/no values or as contribution levels. The grid is the heatmap — transpose it, swap initials for full names, and recolor it in place
- Statements: render by role or by author, with full names or initials and optional level labels
- Localized output: translate role names in statements, Markdown tables, and heatmaps (via credit-translation). Machine-readable exports keep the canonical English CRediT terms
- Heatmap: download the grid as SVG or PNG
- Exports: copy or download JATS4R XML, CSV, JSON, and Markdown
- Sharing & import: encode a draft in a URL, paste names, or import JSON, CSV, or JATS4R XML
| First run | Statement & export |
|---|---|
 |
 |
TypeScript 6 throughout, on pnpm workspaces (app at the root, reusable packages/core).
| Layer | Choice | Why |
|---|---|---|
| Frontend | Next.js 16 (App Router) | Runs on Cloudflare Workers via OpenNext |
| Styling | Tailwind CSS v4 | Design tokens via @theme; no runtime CSS |
| State | Zustand + immer + persist | Survives a refresh via localStorage |
| Validation | Zod | Schema checks at trust boundaries |
| Heatmap | @nivo/heatmap + hand-crafted SVG (core) |
One SVG source feeds both download and canvas PNG |
Browser
└─ Next.js app (repo root, App Router)
├─ React UI + Zustand store (persisted to localStorage)
├─ @credit-generator/core ← all domain logic, runs in the browser
│ statements · JATS4R XML · CSV · JSON · Markdown · heatmap SVG · validation
└─ /api/orcid (route handler) ──→ pub.orcid.org ← the only server-side call
Nearly everything runs in the browser. packages/core holds the domain
logic — statements, exports, validation, XML import (native DOMParser), and the heatmap SVG — as
pure TypeScript with one runtime dependency, zod. The PNG download is drawn from that same SVG onto
a <canvas>.
The one server-side call is the ORCID lookup: the ORCID public API sends no CORS headers, so
/api/orcid proxies it through a small Next.js route handler. (/health backs uptime monitors.)
Contributions are stored as a 0–100 integer score rather than a boolean, so the UI can switch
between binary and level-based editing without changing the stored model. See
packages/core/README.md for the score-to-level boundaries.
Prerequisites: Node ≥ 26, pnpm ≥ 11, just (optional)
git clone https://github.com/simonvanlierde/credit-heatmap
cd credit-heatmap
pnpm install
pnpm dev # → http://localhost:3000See CONTRIBUTING.md for the full command list, dev workflow, and the
lint/typecheck/test checklist. Run just to list the watch/fix recipes layered on the pnpm scripts.
The live demo runs on Cloudflare Workers via
@opennextjs/cloudflare, which adapts the Next.js build.
Pipeline: push to main → Cloudflare
Workers Builds builds and deploys to
credit.duinlab.nl. The trigger is configured in the Cloudflare
dashboard (settings mirrored in wrangler.jsonc); non-main branches and PR
previews are disabled. CI only lints, tests, and build-checks.
Manual deploy (needs Cloudflare credentials):
pnpm preview # build + run the Worker locally
pnpm deploy # build + deploy to CloudflareConfig lives in wrangler.jsonc and open-next.config.ts
- Localize the app UI. Today only the output (statements, Markdown tables, heatmaps) uses the bundled role translations; the interface itself is English-only.
- Widen locale coverage. Only a curated subset of
credit-translation locales is
vendored under
packages/core/src/credit-i18n/translations; refresh them withnode packages/core/scripts/fetch-credit-translations.mjs. - Smooth onboarding. Better empty states and a gentler path from a blank workspace to a finished statement.
Bug reports and small features are welcome, see CONTRIBUTING.md for setup, testing, and the accessibility checks. Design decisions are recorded as ADRs.
The CRediT Generator builds on prior tools and scholarship on contributorship:
- The original Python/Dash CRediT Generator, which inspired this app.
- Role translations from credit-translation.
- The contribution matrix proposed by Nick Steinmetz (2019), the visual form this app's heatmap descends from, as surveyed in Nature Index, "Researchers are embracing visual tools to give fair credit…".
- Brand, A., Allen, L., Altman, M., Hlava, M., & Scott, J. (2015). Beyond authorship: attribution, contribution, collaboration, and credit. Learned Publishing, 28(2), 151–155. https://doi.org/10.1002/leap.1210
- Holcombe, A. O., Kovács, M., Aust, F., & Aczel, B. (2020). Documenting contributions to scholarly articles using CRediT and tenzing. PLOS ONE, 15(12), e0244611. https://doi.org/10.1371/journal.pone.0244611
- Nakagawa, S., Ivimey-Cook, E. R., Grainger, M. J., O'Dea, R. E., et al. (2023). Method Reporting with Initials for Transparency (MeRIT) promotes more granularity and accountability for author contributions. Nature Communications, 14, 1788. https://doi.org/10.1038/s41467-023-37039-1
If you use the CRediT Generator in your work, please cite it. Metadata lives in CITATION.cff, and GitHub's "Cite this repository" button generates APA and BibTeX from it. The archived, versioned release is on Zenodo: doi:10.5281/zenodo.21213659.
van Lierde, S. CRediT Generator [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.21213659
MIT © Simon van Lierde