Run autonomous codebase improvement loops with deterministic benchmark scoring.
npx skills add https://github.com/xdkp/oh-my-claudecode --skill self-improve-xdkp
Please help me install this Agent Skill. Skill: self-improve Source: https://github.com/xdkp/oh-my-claudecode/tree/main/skills/self-improve Command: npx skills add https://github.com/xdkp/oh-my-claudecode --skill self-improve-xdkp
It removes the manual, trial-and-error burden of improving a codebase by running an end-to-end loop that proposes, implements, and objectively benchmarks multiple competing changes until a measurable improvement is found.
Start a Claude Code / OMC session and run the self-improve setup flow so the loop can create a goal, build/confirm a benchmark, and then begin autonomous iterations on your target repository.
High-intent search queries and answers about installing and using this skill.
You prevent regression during automated code optimization by enforcing improvement/hold rules and using sealed-file validation, ensuring the loop cannot modify benchmark evaluation code while confirming no regressions before merging tournament winners.
You prevent regression during automated code optimization by enforcing improvement/hold rules and using sealed-file validation, ensuring the loop cannot modify benchmark evaluation code while confirming no regressions before merging tournament winners.
You prevent regression during automated code optimization by enforcing improvement/hold rules and using sealed-file validation, ensuring the loop cannot modify benchmark evaluation code while confirming no regressions before merging tournament winners.
You prevent regression during automated code optimization by enforcing improvement/hold rules and using sealed-file validation, ensuring the loop cannot modify benchmark evaluation code while confirming no regressions before merging tournament winners.
You prevent regression during automated code optimization by enforcing improvement/hold rules and using sealed-file validation, ensuring the loop cannot modify benchmark evaluation code while confirming no regressions before merging tournament winners.
AlphaCompositeCreate isolated git worktrees sharing one repository for concurrent branch work.
mvillmowRemove merged or stale git worktrees after verifying a clean state.
masharrattCoordinate container-based CFN Loop execution across agents and consensus workflows.
r3dlexAutomate software optimization through tournament selection, benchmarking, and automated merging.
