Pagination

RSC GraphQL list operations use cursor-based pagination. A cursor is a bookmark the server returns that marks where you left off in a result set — you pass it back on the next request and the server picks up from there. This pattern exists instead of page numbers or skip/limit because offset-based pagination gets expensive at scale and produces inconsistent results when data changes between requests.

How it works

List operations return a Connection type. Pass after: null on the first call, then set after to pageInfo.endCursor from each response and repeat while pageInfo.hasNextPage is true.

Cursors are opaque identifiers

Treat them as strings you pass back verbatim, as they may abstract implementation details that will change.

query {
  exampleConnection(after: null) {
    nodes {
      id
      name
    }
    pageInfo {
      endCursor
      hasNextPage
    }
  }
}

Common mistakes

• Don't parse or decode cursor values — their format is not guaranteed.
• Don't cache cursors across sessions or assume they're stable after data changes.
• Always request pageInfo even if you expect a single page — hasNextPage: false confirms a complete result set.