Reference Asset
Topic hubs linking benchmark, dataset, database, and alternatives.
Comparison hubs cluster the catalog by topic using paper-level tags. A topic graduates to a hub once at least three indexed papers tag it; we keep the top 20 by paper count to keep each cluster meaningful.
Each hub joins four asset types in one place — benchmark slice, dataset slice, database query, and the alternatives set — so a reader can navigate the same topic across every reference asset without retyping context.
JSON over HTTPS. Hub-detail endpoints return the cross-asset payload for a single topic.
GET /api/v1/resources/comparisons
GET /api/v1/resources/comparisons/{slug}Use This Via API or MCP
Fetch cross-asset topic payloads in one call. REST and MCP are the only supported transports.