A deliberate first step
The participation portal includes a small synthetic graph-classification exercise. It tests the interface for instructions, consent, storage, feedback and withdrawal. It is not a shrimp observation dataset and produces no scientific findings.
Separating interface trials from research records makes future reporting easier to audit. A useful public contribution needs a question, a reviewer and a clear account of how it will be used.
Towards live tasks
A future study would require licensed data, subject-matter review, quality controls and an agreed retention policy. No timetable or participation total is claimed in this edition.