claude

claude

Global Claude Code configuration. Stows claude/.claude/ → ~/.claude/, which Claude Code reads on every session regardless of cwd.

Install

cd ~/dotfiles && stow claude

Layout

Path	Purpose
CLAUDE.md	Global instructions — how I want Claude to work with me (tone, TS conventions, testing, security)
skills/	Slash-command skills available in any session
agents/	Sub-agent definitions
hooks/	Lifecycle hooks
settings.json	Permissions, model defaults, env vars
statusline-command.sh	Status line script

Skills

Skill	Purpose
/grill-me	Socratic interview to stress-test a plan before implementing
/grill-with-docs	Same as /grill-me, plus captures decisions in repo docs/ADRs
/challenge	Engineering-judgement coaching on architecture decisions
/to-prd	Synthesize current conversation into a PRD on the issue tracker
/to-issues	Break a plan or PRD into independently-grabbable vertical-slice issues
/triage	Move issues through the triage state machine (needs-triage → ready-for-agent/ready-for-human/wontfix)
/tdd	Red-green-refactor loop with the agent
/diagnose	Disciplined diagnosis loop for hard bugs and perf regressions
/improve-codebase-architecture	Find deepening opportunities informed by domain language and ADRs
/review	Review code changes for quality, conventions, security, test coverage
/commit	Conventional commit, single-line message, no co-authoring
/pr	Open a pull request with a structured summary
/validate	Run typecheck, lint, and tests in one go
/write-skill	Scaffold a new skill with progressive disclosure
/obsidian-vault	Search, create, manage notes in the Obsidian vault

Pipeline

The skills compose into a deliberate workflow:

ideation → /grill-me or /grill-with-docs
        → /to-prd        (parent issue, needs-triage)
        → /to-issues     (vertical slices, needs-triage)
        → /triage        (ready-for-agent | ready-for-human)
        → ralph          (AFK execution — see ../ralph/)

See ralph for the AFK execution layer.