FDP Ecosystem Template

Start your own static FAIR Data Point ecosystem: Turtle metadata in a GitHub repository, served by GitHub Pages, with a browsable website, self-service contributions and automatic FAIR assessment. No server needed.

Each catalog describes one resource and every way to obtain it: portal, APIs, SPARQL endpoints, bulk dumps, cloud buckets, copies in aggregators. Each way is typed with an access method and comes with a "How to get it" recipe.

Start here: https://fdp.semscape.org/ecosystems/ (the landing page, with all ecosystems and a one-click start). Demo: https://fdp.semscape.org/fdp-ecosystem-template/ (this repository, with one example source). In use: Biodiversity FDP (source).

Start your own

1. Click Use this template → Create a new repository. Keep it public: GitHub Actions and Pages are then free.

2. In the new repository, go to Settings → Pages and set Source to GitHub Actions.

3. Go to Actions → Deploy viewer to GitHub Pages → Run workflow. The first deploy initialises your copy:

   • points fdp.config.json at your repository and site
   • rebases every IRI
   • writes your README.md and PERSISTENCE.md

   Your site appears at https://<owner>.github.io/<repository>/.

4. Make it yours:

   • fdp.config.json: name, headline (*word* is emphasised), description, keywords, maintainer, colours, fonts
   • assets/logo.svg: your logo, also used as the favicon. It is drawn in the theme colour wherever it uses currentColor.

5. Replace the example: remove fdp/wikidata-fdp/ and its entry in fdp/index/catalog.ttl. Then add sources through Issues → New issue → Add a data source, or follow CONTRIBUTING.md.

Optional:

• Let the bot open pull requests: Settings → Actions → General → Allow GitHub Actions to create and approve pull requests. This is used by the issue form and the engine updates.
• Persistent identifiers: register a w3id, then run scripts/set-base.sh https://w3id.org/<your-id>/.

What you get

Website	Index → catalog → dataset → "ways to get it", search, access-method filters, contributors, FAIR results; light and dark mode; static HTML and schema.org JSON-LD for search engines
Live FDPs	An index entry can point at someone else's FAIR Data Point. The site crawls it in the browser.
Self-service	An issue form becomes a validated pull request (scripts/add_source.py)
Credit	Contributors from commits, pull requests and form submissions, with names, Wikidata items and ORCID iDs from Wikidata (P2037). Published in the site and as RDF (dcterms:contributor).
FAIR assessment	Walks every level of the FDP and runs the OSTrails FAIR Champion tests: weekly, after deploys and on demand
Validation	Every push checks that the Turtle parses, the DCAT shape, the vocabulary terms and the index consistency

Engine and content

Yours (content)	The engine (from this template)
fdp.config.json, fdp/, assets/, README.md, PERSISTENCE.md	scripts/, site/, .github/, fair/tests.txt, CONTRIBUTING.md, CLAUDE.md. The full list is in .fdp-engine.

Engine files read everything instance-specific from fdp.config.json (via scripts/fdpconfig.py), so the engine can be updated without touching your content:

• Automatically: Actions → Update engine (also runs monthly) opens a pull request with the template's latest engine.

  GitHub's built-in token cannot change workflow files. Add a repository secret ENGINE_UPDATE_TOKEN (a fine-grained token with Contents, Pull requests and Workflows write access) to include them; otherwise the pull request lists the workflow changes for you.

• Locally: scripts/update-engine.sh, then review with git diff and commit.

Local preview

pip install rdflib
python3 scripts/validate.py
scripts/build-site.sh && python3 -m http.server -d _site 8000

License

The code is released under the MIT license. The FDP metadata is released under CC0 (see the index catalog).