Post #9824

Innei 向 mx-space/core 提交了一个更改 feat(ai): implement Vercel AI Gateway prompt caching in OpenAICompatibleRuntime - Added baseURL detection for Vercel AI Gateway to enable automatic caching for chat completion requests. - Introduced helper methods to decorate request parameters with `providerOptions.gateway.caching = 'auto'` for compatible endpoints. - Updated `generateText`, `generateStructured`, and `generateTextStream` methods to utilize the new caching logic. - Created unit tests to verify caching behavior for both gateway and non-gateway endpoints. Signed-off-by: Innei <[email protected]>