本文へ移動
cccskills
無料GitHub で公開

ralph-specum-help

This skill should be used only when the user explicitly asks to use `$ralph-specum-help`, or explicitly asks Ralph Specum in Codex for help or command guidance.

インストール方法を見る

含まれるファイル(2)

  • SKILL.md3.5 KB
  • agents/openai.yaml268 B

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Ralph Specum Help

Use this to explain the Ralph Specum surface in Codex.

Cover

  • Primary skill: $ralph-specum
  • Helper skills: $ralph-specum-start, $ralph-specum-triage, $ralph-specum-research, $ralph-specum-requirements, $ralph-specum-design, $ralph-specum-tasks, $ralph-specum-prototype, $ralph-specum-implement, $ralph-specum-status, $ralph-specum-switch, $ralph-specum-cancel, $ralph-specum-index, $ralph-specum-refactor, $ralph-specum-feedback, $ralph-specum-help
  • Normal flow: start goal grill, approval, research, artifact approval, requirements grill, approval, design grill, approval, tasks grill, approval, implement
  • Large effort flow: triage, then start each unblocked spec
  • Quick mode: only exact --quick bypasses interview questions; exact --interactive clears it. Discovery, skill loads, and artifact write gates still run
  • Disk contract: ./specs or configured roots, .current-spec, optional .current-epic, per-spec markdown files, .ralph-state.json

Optional Prototype Overlay

  • Direct mode: invoke $ralph-specum-prototype from research, requirements, design, tasks, or execution. Preserve the main phase and current checkout.
  • Suggested mode: research or requirements may offer continue to prototype, then return to the next normal phase.
  • Resume mode: use $ralph-specum-prototype --resume <id>. One active entry resumes automatically; several are listed deterministically for selection.
  • Quick mode: run exactly one agent-owned request after requirements, ask no prototype questions, take over the oldest design blocker or select the highest-risk grounded, falsifiable question when no blocker exists, own verdict and handoff decisions, and continue to design for every result.
  • Cancel mode: use $ralph-specum-prototype --cancel <id> or safe spec cancellation. Stop at a safe boundary, publish a reviewed immutable cancelled record, and preserve source and partial implementation.

Store terminal records under the resolved <basePath>/prototypes/. Run source work in a sibling worktree or eligible scratch area without switching the current checkout or copying unapproved dirty paths. Keep source, evidence, records, and branches local. Require separate authorization for every push, PR or issue change, and other remote action. Require exact-path and local-branch confirmation before local deletion; never delete a remote branch during prototype cleanup.

Guidance

  • Recommend $ralph-specum as the default entrypoint.
  • Recommend $ralph-specum-triage when the user describes a large, multi-part, or dependency-heavy effort.
  • Mention helper skills when the user wants explicit phase control.
  • Explain that Ralph does not self-advance by default. The user must approve the current artifact, request changes, or explicitly continue to the next step.
  • Explain that safe cancel deletes execution state only after verified immutable cancellation records exist. Full spec deletion and each prototype source deletion are separate exact-target confirmations.
  • Explain that $ralph-specum-status reports active entries, candidates, immutable finals, quarantines, blockers, return phase/task, and source disposition.
  • Mention optional bootstrap assets only when the user wants repo-local guidance.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.

日本語の概要は準備中です。原文の説明を表示しています。

tzachbon/smart-ralph5592026年9月16日 更新

This skill should be used when generating spec artifacts (research.md, requirements.md, design.md, tasks.md), formatting agent output, structuring phase results, or when any Ralph agent needs guidance on concise, scannable output formatting. Applies to all Ralph spec phase agents.

日本語の概要は準備中です。原文の説明を表示しています。

tzachbon/smart-ralph5592026年9月16日 更新

Output rules for all agents - concise, scannable, actionable. Based on Matt Pocock's planning principles.

日本語の概要は準備中です。原文の説明を表示しています。

tzachbon/smart-ralph5592026年9月16日 更新

Core principle that the main agent is a coordinator, not an implementer. All work must be delegated to subagents.

日本語の概要は準備中です。原文の説明を表示しています。

tzachbon/smart-ralph5592026年9月16日 更新

This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.

日本語の概要は準備中です。原文の説明を表示しています。

tzachbon/smart-ralph5592026年9月16日 更新

This skill should be used when a Ralph phase must identify critical user decisions, run a layered grill, persist partial answers, obtain explicit approval, or resume an interrupted phase interview before delegating artifact work.

日本語の概要は準備中です。原文の説明を表示しています。

tzachbon/smart-ralph5592026年9月16日 更新

tzachbon のスキルをすべて見る

このスキルの問題を報告する