test(ci): verify F-UJI CI end-to-end

[andreifoldes]

Synthetic same-repo PR to verify the new two-stage F-UJI CI works end-to-end. Adds a single <link rel="cite-as"> Signposting header to dataset/index.html — expected to nudge the F-UJI score upward.

Will close once the CI green-lights the rendered check + comment.

[github-actions]

🧪 F-UJI FAIRness assessment

FAIR: 8.33% (was 8.33% on main — · 0.00)

Dimension	main	PR	Δ %	Maturity
F	1/7 (14.29%)	1/7 (14.29%)	· 0.00	1 → 1
A	1/3 (33.33%)	1/3 (33.33%)	· 0.00	1 → 1
I	0/4 (0.00%)	0/4 (0.00%)	· 0.00	0 → 0
R	0/10 (0.00%)	0/10 (0.00%)	· 0.00	0 → 0

_{Scored by F-UJI 3.5.0 against http://host.docker.internal:8000/. Full JSON in the workflow artifact.}

[andreifoldes]

Close+reopen to force fresh pull_request event — synchronize events on the last two pushes mysteriously didn't fire F-UJI Score.

[andreifoldes]

🧪 F-UJI scoring — pipeline fix summary

The F-UJI Score CI is now fixed and pinned: every PR (open & closed) is scored against a single fixed baseline = the repo's initial commit 75466a8, so deltas are comparable across the whole hackathon.

This PR's score: FAIR 8.33%  ·  Δ +0.00 vs initial

Main reason: Only a <link rel="cite-as"> Signposting header was added. Signposting links alone don't carry license / descriptive metadata — F-UJI needs an extractable metadata document (metadata.jsonld or DCAT metadata.xml).

---

For the full scoreboard across all 11 PRs and the patterns that move F-UJI's needle, see the project notes. The CI will keep this score up to date on any new pushes; conflicted/closed PRs were scored manually using the same pipeline against the same baseline.