skills/src/setup-ralph/SKILL.md
Prepares .ralph/ runtime state from an approved and reviewed plan bundle. Syncs plan tasks into prd.json and updates prompt.run.md with project-specific context and quality gates. Use after plan approval, decompose-plan review PASS, and ralph init.
npx skillsauth add shuymn/dotfiles setup-ralphInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
<skill-root> means the directory containing this SKILL.md.scripts/... and references/... relative to <skill-root>, not the caller's current working directory.Prepare .ralph/ for the execution loop: fill prd.json from an approved plan bundle and update prompt.run.md with project-specific guidance.
Prerequisite: .ralph/ has already been initialized via ralph init.
Do NOT sync a plan that has not been explicitly approved by the user.
plan.md path from $ARGUMENTS or docs/plans/*/plan.md.skit bundle-validate-check <plan-file>.Checkpoint Summary contains:
Alignment Verdict: PASSScope Contract Guard: PASSQuality Gate Guard: PASSReview ArtifactTrace PackCompose PackUpdated Atplan.review.md from Review Artifact or the plan location.skit gate-check <review-file> <plan-file>.Overall Verdict: PASSTask Shape Blockers: noneScope: PASSExecution Readiness: PASSExtract each ### Task N: <title> section and sync it to one Ralph story.
Required story metadata:
id: task-Ntitledepspasses: falserisk_tierscope_contractboundary_requiredcompletion_gateRules:
scope_contract must preserve Owned Paths, Shared Touchpoints, and Prohibited Paths.boundary_required is true when the task declares Boundary Verification.completion_gate is true when the plan as a whole requires completion-audit, or when the task belongs to a public/runtime/release-claim path that cannot be closed by local task PASS alone..ralph/prd.jsonOverwrite .ralph/prd.json with:
{
"project": "<derived-project-slug>",
"plan": "<relative-path-to-plan.md>",
"branchName": "<derived-branch-name>",
"stories": [
{
"id": "task-1",
"title": "<task title>",
"deps": [],
"passes": false,
"risk_tier": "Standard",
"scope_contract": {
"owned_paths": ["path/to/owned-area/**"],
"shared_touchpoints": [
{
"path": "path/to/shared-file.ext",
"rationale": "shared configuration or dependency update"
}
],
"prohibited_paths": ["path/to/prohibited-area/**"]
},
"boundary_required": false,
"completion_gate": false
}
]
}
.ralph/prompt.run.mdUpdate only the editable project-specific sections:
AGENTS.mdScope Contract handling guidanceexecute-plan for each storycompletion-audit when the plan or story metadata marks a completion gate## Quality GatesSummarize:
completion-audit is requiredStop and ask the user when:
.ralph/ does not exist.ralph/prd.json or .ralph/prompt.run.md does not existCheckpoint Summary or required PASS guardsplan.review.md is missing or not PASSTask Shape Blockers is not noneAGENTS.md or CLAUDE.md) existsprd.json is a projection of plan.md; do not invent or reorder stories.Scope Contract metadata; do not collapse it into flat file lists.completion-audit, keep that signal in Ralph state.prompt.run.md.development
Simplifies recently changed code by running three parallel reviews (reuse, quality, efficiency) and applying only behavior-preserving fixes. Use when the user asks to simplify, clean up, reduce duplication, improve code reuse, or optimize recently changed code, a staged diff, a branch diff, or explicitly listed files. Also use when the user says things like 'simplify this', 'まとめて整理して', 'コードをスリムにして', or invokes `/simplify`.
tools
Use when the user invokes /workflow. Injects project workflow methodology as context. Accepts argument: plan, exec, review (default: all).
development
Processes AI reviewer feedback and applies only verified fixes. Works in two modes: (1) fetches comments from a PR URL or current branch, (2) processes feedback pasted directly into the conversation. Trigger when the user wants to bulk-process or apply AI review suggestions — from a GitHub PR or pasted text. Do NOT trigger for single questions about what a bot said, or general code review discussion.
development
Review text for AI writing tropes and mechanical prose patterns, then report findings with concrete rewrite suggestions. Use when the user asks to proofread, lint, or polish writing — especially docs, blog posts, READMEs, commit messages, PR descriptions, or any prose that should read as human-written. Also use when the user says "check for AI slop", "does this sound like AI", "make this sound more natural", or wants to de-AI their text.