ワンクリックで
dotfiles
dotfiles には Givemegitpls から収集した 11 個の skills があり、リポジトリ単位の職業カバレッジとサイト内 skill 詳細ページを表示します。
このリポジトリの skills
Use when migrating a Python project from Poetry to uv. Covers running `uvx migrate-to-uv`, fixing pyproject.toml conversion gaps, rewriting Dockerfiles (poetry install -> uv sync), updating .gitlab-ci.yml, and verifying the result. Trigger keywords: migrate poetry to uv, migrate-to-uv, poetry -> uv, switch to uv, replace poetry.
Use when committing, creating PRs, or managing git history. Conventional commits, pre-commit hooks with ruff and basedpyright, branching conventions.
Use when first opening an unfamiliar Python project. Quick analysis of project structure, dependencies, entry points, test setup, and tooling.
Use when working with poorly typed or legacy Python code. Covers gradual typing, safe annotation strategies, dynamic attributes, and monkeypatching without breaking behavior.
Use when refactoring Python code safely. Covers extract method/class/module, dependency inversion, constants extraction, and test-protected changes.
Use when writing or editing Python code. Enforces strict typing with basedpyright, ruff linting and formatting, modern Python 3.12+ idioms.
Use when writing tests for Python code. Covers pytest conventions, fixtures, mocking, strict type-checked tests.
Use when building, compiling, or troubleshooting Minecraft Paper/Spigot/Purpur plugins. Covers Gradle builds, Java toolchain setup, Paper API, plugin.yml, JAR assembly, and common build failures.
Use when listing, extracting, or creating archives (.zip, .jar, .7z, .tar.gz, etc.) and unzip/jar are unavailable. Covers 7z list, extract, and add commands with common switches.
Use when creating or modifying GitLab CI pipelines for APS projects. Covers CI components registry, image-build, release, docs, and pipeline conventions.
Use when writing or editing a SKILL.md, agent system_prompt, command template, or docs that reference CLI commands (uv, poetry, ruff, git, docker, etc.). Enforces verifying every command example and flag against actual --help or man output on the target machine before writing it. Also use when a skill starts producing wrong commands.