4.0.0-rc.1+20260812

Interpretation limitations

Network proximity and centrality do not measure scientific quality, leadership, or contribution. Algorithmic communities are working hypotheses, not automatically established scientific schools.

• A DOI search boundary or provider failure is not proof that a DOI does not exist.
• Metadata-only access does not imply absence of a full text elsewhere.
• Author order is not evidence of contribution, supervision or leadership.
• A person can belong to multiple groups; no primary group is forced.
• Missing translations are warnings and display their fallback status.
• Centrality is descriptive and must not be used for researcher rankings.
• The current release visualizes co-authorship, authorship, component work and organization affiliation. Supervision and shared-project relations exist in person_relations but are not yet included in graph projections.

Release warnings

• Preview v4 release; it is not a stable data activation.
• The project owner confirmed these 20 records only as the latest publicly documented component-line contexts on 2026-08-12.
• A context does not establish current implementation or software/data availability, implementation quality, scientific validity, or field suitability.
• The selected evidence remains bibliographic publication evidence; it does not independently validate the model or implementation.
• The five components remain project-authored corpus views, not a universal AgroTool ontology.
• Spanish and Chinese context translations are not approved and use visible Russian fallback behavior.
• Owner D-48/D-49 interface acceptance remains pending until this data candidate is deployed as an exact beta build.