원클릭으로
refactor-claude-md
Use when refactoring a user-level or project-level CLAUDE.md for progressive disclosure
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Use when refactoring a user-level or project-level CLAUDE.md for progressive disclosure
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use when brainstorming ideas, features, or directions for a project where independent perspectives from different model families (Claude/Codex/Gemini) would surface blind spots and spark creative options the user hasn't considered — especially "what cool things can I add", "what should I build next", "give me ideas for X"
Use when refactoring a user-level or project-level AGENTS.md for progressive disclosure
Use when creating or editing a skill that uses Bash commands, external tools, or skill invocations and the allowed-tools frontmatter may be incomplete
(project) Use when bumping hardcoded tool versions in playbooks/roles/ to their latest releases
(project) Use when editing any file under skills/hal-skills/ or plugins/hal-voice/ to bump the plugin version before committing
Use when making any git commit. Always pass a brief description of what changed as the argument.
| name | refactor-claude-md |
| description | Use when refactoring a user-level or project-level CLAUDE.md for progressive disclosure |
| argument-hint | ["user | project"] |
| user-invocable | true |
| disable-model-invocation | true |
| model | fable |
| allowed-tools | ["AskUserQuestion","WebFetch","Edit(~/.claude/CLAUDE.md)","Edit(~/.claude/rules/**)","Edit(CLAUDE.md)","Edit(.claude/rules/**)"] |
Refactor a CLAUDE.md toward progressive disclosure. Every line it keeps is loaded into every session the file covers, and a frontier model follows only around 150 to 200 standing instructions consistently, so each line competes for that budget. Everything else moves down the ladder or out of the file.
Pick the target. If the invocation names one, use it. Otherwise ask with AskUserQuestion, one option per file that exists, with the resolved path in the label:
~/.claude/CLAUDE.mdFetch the yardsticks. Fetch these pages. They calibrate the delete and rewrite verdicts below:
Audit. Read the file and give every instruction exactly one verdict, judged against the target's bar below. Done when no instruction lacks one.
The bar per target:
paths:-scoped file in ~/.claude/rules/.paths:-scoped file in .claude/rules/, or a doc file linked from CLAUDE.md.Get decisions. One AskUserQuestion per contradiction, with each conflicting version as an option. Then present the delete and demote lists for sign-off. The file stays untouched until sign-off.
Rewrite. Apply the signed-off verdicts in one pass. When creating a new rules file, fetch https://code.claude.com/docs/en/memory#path-specific-rules for the paths: frontmatter syntax. Done when every audited instruction landed where its verdict says: kept, rewritten, demoted, or deleted.