Enable <XML> Tag Based UI Extraction for Action Generation which is only available for reasoning( extractReasoningMiddleware)

[shadee22]

I want to enable tag-based UI extraction from the stream, similar to how extractReasoningMiddleware extracts reasoning sections. Specifically, I aim to generate actions using a response stream and display them properly in the UI as separate "actions" elements. Using the Vercel AI SDK, what is the best approach to handle this? The middleware should extract content between custom tags,

so using this kind of approach i need to show actions after that response. for reasoning i can do it by type.reasoning and do this. but for actions how to do that.

[github-actions[bot]]

This discussion was automatically locked because it has not been updated in over 30 days. If you still have questions about this topic, please ask us at community.vercel.com/ai-sdk