Context7 makes its search order easier to keep intact.
The 0.6.0 SDK update types a shared rank field across code and information snippets. A small change for people building their own documentation tools.

The update
The @upstash/context7-sdk@0.6.0 release, published on 23 October, says search results now type the rank field on each code and information snippet. The release defines that value as the snippet’s one-based position in a reranked order shared by both lists.
What that means in practice
A developer can receive code examples in one list and explanatory information in another. If an interface simply prints one whole list before the other, it may lose the combined order. The release note says to sort the two kinds of snippet together by rank to reproduce the order used by the text response.
Who should pay attention
This is most relevant to someone using the SDK to build a custom search view, an agent tool or a documentation interface. It is not, by itself, a new model or evidence that an agent will stop making mistakes. If you only use an existing integration, the detail may be handled by the tool around it.
What we have not verified
This write-up is based on the project’s release note. We have not run the new SDK, compared search quality or measured the effect on an agent’s answers. We are describing a typed ordering field, not claiming that the search algorithm has been replaced.
Our read
The useful detail is the shared order. When an interface combines several result types, preserving the source’s ordering is part of presenting those results faithfully. For builders already consuming the JSON response, it is a focused change to inspect in their own code.