self-improve

You can automate code improvement by using tournament selection to compare code changes and benchmarking to validate them. This Skill executes that process, requiring various OMC agents like the planner, architect, and executor to analyze, plan, and apply the optimizations.

Tournament selection in code improvement is a mechanism that compares different code changes against each other, evaluating performance to select and integrate the best-performing version. This Skill uses it to ensure continuous integration yields measurable optimization results.

To benchmark code changes, this Skill compares variations using tournament selection and validates performance before applying them. It leverages OMC agents including the critic and executor to analyze code and execute the validated improvements into the target codebase.

Yes, this Skill requires multiple OMC agents including the planner, architect, critic, executor, and git-master, plus opus. These dependencies are necessary to perform the code analysis, planning, execution, and git operations required for continuous improvement.

Yes, this automated optimization applies to complex software development projects. It analyzes code to automatically improve performance, maintainability, and readability, allowing developers to focus on more important tasks while the tournament selection handles validation.

A key limitation is that the automated optimization process strictly requires multiple OMC agents and opus to function. Without these specific dependencies for code analysis and git operations, the tournament selection and benchmarking workflow cannot execute.