This pull request introduces improved handling of API error responses across both backend and frontend, with a focus on better error messaging, quota/rate limit awareness, and user experience. The backend now passes a pro flag to request control logic, ensuring that Pro users bypass quota checks. On the frontend, a new unified error parsing utility provides clearer feedback to users, especially around quota and rate limiting, and controls when retries are allowed.

The most important changes are:

Backend: Pro User Handling and Quota Checks

• Added an is_pro flag to the enforce_request_controls function and its callers, ensuring that Pro users are exempt from daily quota checks. This affects both chat message and topic query endpoints. [1] [2] [3] [4] [5] [6]

• Improved streaming response handling to ensure that partial responses and errors are properly cached and labeled, and that users are informed when a connection is interrupted, with messaging tailored to the selected mode (e.g., "technical" or "socratic"). [1] [2]

Frontend: Unified API Error Handling and Messaging

• Introduced a new buildApiError utility in src/lib/httpErrors.ts to standardize error parsing and messaging for API responses, including quota and rate limit errors. [1] [2] [3] [4]

• Updated all API request functions (fetchAPI, queryTopic, queryTopicStream, and chat service) to use the new error utility, providing more detailed and user-friendly error messages and controlling retry logic based on error detail (e.g., not retrying on quota exceeded). [1] [2] [3] [4] [5] [6] [7]

API Improvements

• Enhanced timeout and abort handling in API requests to provide more accurate error messages when requests time out or are aborted, and to prevent retries when not appropriate. [1] [2] [3] [4] [5]

These changes result in a more robust user experience, clearer error feedback, and improved handling of Pro user privileges and quota enforcement.