Dot Claude Code

dot-claude-code

 

Personal Claude Code config — rules, skills, agents, and hooks versioned in git and symlinked into `~/.claude/`.

Structure

global/
├── CLAUDE.md              # Core instructions
├── settings.json          # Permissions, hooks, feature flags
├── statusline-command.sh  # Status line script
├── rules/                 # Auto-loaded topic rules
├── skills/                # Slash command workflows
├── agents/                # Subagent definitions
└── hooks/                 # Pre/PostToolUse shell hooks
templates/
└── skill-template/        # Starter for new skills
docs/
└── agentic-framework/     # Agent/workflow design model

Rules

File	Covers
git.md	Conventional commit format
working-approach.md	Validate before reporting done
verify-before-assume.md	Verify external facts before using them

Skills

Skill	Command	What it does
claude-md-audit	/claude-md-audit	Audit and rewrite CLAUDE.md files
conclude	/conclude	Pre-close session review for loose ends
diagnose	/diagnose	Disciplined loop for hard bugs and perf regressions
format-amznstyle	/format-amznstyle	Flag weasel words/jargon in writing, Amazon style
grilling	/grilling	Relentless interview to stress-test a plan
grill-me	/grill-me	Thin wrapper around grilling
skill-creator	/skill-creator	Create and audit skills with quality scoring
skill-refine	/skill-refine	Apply learnings from a skill run back into its SKILL.md
validate	/validate	Mid-session READY/NEEDS WORK quality ga