Identify dead code, unused imports, and duplicated logic in ML experiment scripts.
npx skills add https://github.com/daikichiba9511/dotfiles --skill ml-cleanup
Please help me install this Agent Skill. Skill: ml-cleanup Source: https://github.com/daikichiba9511/dotfiles/tree/main/private_dot_codex/skills/ml-cleanup Command: npx skills add https://github.com/daikichiba9511/dotfiles --skill ml-cleanup
ML project codebases accumulate dead code and duplication that waste time and hinder reproducibility.
Run ml-cleanup on an experiment directory to identify dead code and generate a cleanup plan.
High-intent search queries and answers about installing and using this skill.
To remove dead code from Python ML experiment scripts, apply a cleanup process to your experiment directory to identify unused imports, duplicated logic, and generate a structured cleanup plan with optional automated patches.
Refactoring duplicated logic across ML experiments involves identifying shared code and extracting it to common modules, reducing maintenance costs while ensuring functionality is preserved and the workflow remains reproducible.
Reducing code bloat in ML projects safely requires generating a concise cleanup plan that targets dead code and unused imports, optionally applying automated patches to ensure validation is robust and functionality is preserved.
Yes, you can automatically patch unused imports in Python ML codebases by running a targeted cleanup operation on an experiment directory, which produces optional automated patches to trim the code safely.
Yes, ML cleanup can be applied to a target directory such as src/exp/{exp_dir} or any specific experiment folder to identify dead code, unused imports, and duplicated logic for refactoring.
Cleaning up shared logic improves ML reproducibility by extracting duplicated code into common modules, reducing maintenance overhead and ensuring the experiment workflow remains robust and consistent.
digital-stoic-orgAutomate monorepo-aware dependency installation across Python, JavaScript, and system packages.
KentoShimizuEnforce ML experiment reproducibility by logging code, data, and configuration versions.
causify-aiRename unused Python functions and classes with leading underscores.
noartemRefactor Laravel controllers into Form Requests, Actions, and DTOs.
