Skill Claude CodeCodex
A Japanese-language skill for summarising an AI agent's work log with scripts and extracting reusable recommendations. It records those recommendations in agent notes only after the user approves them.
Skill Claude CodeCodex
A Japanese-language skill for summarising an AI agent's work log with scripts and extracting reusable recommendations. It records those recommendations in agent notes only after the user approves them.
Skill Claude CodeCodex
A Japanese-language review step that asks questions about a finished result's behavior and purpose. It focuses on whether the user understands and agrees with the outcome, rather than on implementation details.
Skill Claude CodeCodex
A set of instructions for removing local and remote Git branches that have been merged, while keeping the base branch, active work, and branches used by other worktrees. Git is a tool for tracking code changes, and a worktree is another working folder linked to the same repository.
Skill Claude CodeCodex
A guide for reviewing a code change using the project's Git history, including the files changed, a minimum review threshold, three analyses, and a hotspot score based on change frequency and line count.
Skill Claude CodeCodex
A Japanese-language code-review skill that judges implementation quality and whether the code matches its intended change. It uses code smells, project conventions, testability, shallow-module detection, and other review criteria.
Skill Claude CodeCodex
A static code-review skill that uses tools to find problems that can be detected mechanically and reports their output without adding AI judgment.
Skill Claude CodeCodex
A Git workflow that plans how changes should be divided into commits and proposes useful commit messages. Git commits are saved snapshots of project changes.
Skill Claude CodeCodex
A critical review step that tests unfinished work by assuming it may fail. It lists problems with supporting reasons and severity, while leaving the decision to accept or reject them to a person.
Skill Claude CodeCodex
A skill that retrieves implementation details from Figma, a tool for designing digital interfaces, and maps them to a project's CSS framework.
Skill Claude CodeCodex
A design-recreation tool that turns the output of x-figma-extract into individual interface components. It does not design the application architecture, manage state, handle page navigation, or fetch data.
Skill Claude CodeCodex
A skill entry point for starting a grilling-style question process. The exact questioning and answer format come from a separate x-grilling skill.
Skill Claude CodeCodex
A goal-definition interview for work that has not started yet. It asks questions until the purpose, completion conditions, constraints, and out-of-scope work are agreed.
Skill Claude CodeCodex
A skill for asking up to three plain-text questions at a time until a plan, design, or idea is fully specified.
Skill Claude CodeCodex
A paragraph-by-paragraph review process for Japanese documents that identifies claims, supporting reasons, connections, and opposing views. It is designed to clarify reasoning, not to guess what the writer intended.
Skill Claude CodeCodex
A Japanese-language writing review skill for Markdown files. It checks formatting, document structure, and Japanese expression, then can revise those areas in repeated review rounds.
Skill Claude CodeCodex
A workflow for rebasing a feature branch onto a base branch such as main, master, or develop. Rebasing replays your branch's commits on top of the latest commits from another branch.
Skill Claude CodeCodex
A learning and documentation aid that investigates code, documentation, and architecture, then explains the findings step by step in a Markdown file.
Skill Claude CodeCodex
A review guide for improving existing automated tests, including tests in TypeScript files and documentation. TDD means test-driven development, but this add-on is for reviewing tests that already exist rather than writing new ones.
Cursor rule
A set of rules written in Japanese for keeping users informed and helping them make accountable, evidence-based decisions. It says the AI should provide information and prepare the environment so the user can focus on thinking.