Developed OpenAI API usage cost tracking for the posthog-python repository, enabling automated capture of pre-calculated costs from OpenAI Agents responses. Enhanced the backend by updating the tracing processor to extract cost data from the usage object and attach it to trace properties, supporting improved cost visibility and data accuracy. Implemented comprehensive unit tests to verify correct behavior when cost information is present or absent, ensuring reliability and preventing regressions. Utilized Python for backend development and API integration, with a focus on maintainability through minor refactoring and patch generation. This work supports budgeting, optimization, and observability standards within the codebase.