During May 2025, J. Coens developed the Configurable Reasoning Effort Level feature for the openai/codex repository, enabling users to adjust the model’s reasoning depth via a new --reasoning-effort CLI flag. This addition allows for dynamic tuning between response quality and latency, addressing diverse workload and customer requirements. Coens implemented the feature end-to-end in Rust, focusing on maintainable CLI and configuration architecture while leveraging Git for version-controlled delivery. The work demonstrated proficiency in backend development and command-line interface design, resulting in a flexible solution that enhances user control over model behavior without introducing bugs or regressions during the release.