wenjin272 commented on issue #912: URL: https://github.com/apache/flink-agents/issues/912#issuecomment-5707691156
For the second question, I think a parse failure should consume the existing retry budget, consistent with the current behavior. The retry only repeats the current chat attempt with the same message history; it does not rerun completed tool-call rounds. We could optimize this later to retry only the native finalization call, but I do not think that is necessary for the initial implementation. For the third question, I suggest using the same directive in Java and Python: > Convert the previous assistant response into the required structured output format. Preserve its meaning and do not add or infer any new information. The provider already receives the schema separately, so the directive does not need to repeat it. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
