codex-delegator

The Codex Delegator skill facilitates seamless collaboration between Claude Code and OpenAI Codex by automating the delegation of complex tasks. This skill solves the problem of handling intricate tasks that require advanced logical reasoning, backend processes, and problem-solving approaches. When Claude Code identifies tasks that are resistant to conventional methods, it can delegate these to Codex for efficient resolution, ensuring better task execution and optimization. The process involves analyzing the task complexity, invoking Codex to execute the task, and then reviewing the solution for accuracy and completeness before returning the solution to the user. This skill significantly improves the handling of backend logic, complex algorithms, and persistent bugs that require new insights or perspectives to solve effectively.

Main features of the Codex Delegator skill include the automatic analysis and decision-making for task delegation, providing a transparent process where Claude Code informs the user when Codex is invoked. It supports complex backend logic, intricate data transformations, and logic-intensive problems, ensuring that tasks are delegated to Codex only when necessary, optimizing performance and improving task outcomes. The integration with Codex allows for fresh perspectives on recurring issues, such as unsolved bugs or performance bottlenecks, with full context provided to Codex to improve the chances of success.

This skill is ideal for developers and teams who are dealing with complex backend systems, multi-layered architectures, and tasks that require specialized problem-solving. It is especially beneficial for those facing persistent, unsolved issues or who need alternative approaches to standard coding solutions. By using Codex for complex tasks, developers can improve their productivity, streamline problem-solving processes, and ensure that their systems are more efficient and reliable.