本文へ移動
cccskills
無料GitHub で公開

smart-ralph

This skill should be used when the user asks about "ralph arguments", "quick mode", "commit spec", "max iterations", "ralph state file", "prototype overlay", "execution modes", "ralph loop", "coordinator behavior", "delegate to subagent", or needs guidance on Ralph plugin arguments, state management, delegation patterns, or execution loop behavior. Core behavioral skill for all Ralph Specum operations.

インストール方法を見る

含まれるファイル(2)

  • SKILL.md7.6 KB
  • references/state-file-schema.md5.3 KB

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。

Smart Ralph

Core skill for all Ralph plugins. Defines common arguments, execution modes, shared behaviors, and coordinator delegation rules.

Common Arguments

All Ralph commands support these standard arguments:

ArgumentShortDescriptionDefault
--quickExact token that enables persistent non-interactive generationfalse
--interactiveExact token that clears persistent quick mode-
--commit-cCommit spec/feature files locally after generationtrue (normal), false (quick)
--no-commitExplicitly disable committing files-
--max-task-iterations-mMax retries per failed task before stopping5
--fresh-fForce new spec/feature, overwrite if existsfalse

Argument precedence: --no-commit > --commit > mode default.

Execution Modes

Normal Mode (Interactive)

  • Exact --interactive clears a prior authorized quick mode
  • No mode flag resets legacy quick state that lacks exact --quick authorization
  • User reviews artifacts between phases
  • Phase transitions require explicit commands
  • Each phase sets awaitingApproval: true
  • Commits spec files by default
  • Research and requirements may offer the optional prototype overlay
  • The user owns prototype capture, verdict, handoff, and deletion decisions

Quick Mode (--quick)

  • Only an exact --quick argument token enables this mode
  • -q, settings defaults, substrings, and natural-language requests do not enable it
  • --quick with --interactive is an error
  • Skip all interactive prompts, interviews, and approval pauses
  • Run the same phase agents (research, requirements, design, tasks) sequentially
  • Agents receive a "be more opinionated" directive since there is no user feedback
  • spec-reviewer validates each artifact (max 3 iterations)
  • Run at most one agent-owned prototype request after requirements and before design
  • Take over the oldest design-blocking prototype when one exists; otherwise choose the highest-risk grounded question
  • Ask no prototype question or decision and continue to design after every outcome
  • Immediately start execution after all phases complete
  • Do NOT commit by default (use --commit to override)
  • Still delegate to subagents (delegation is mandatory)

State File

All Ralph plugins use .ralph-state.json for execution state. See references/state-file-schema.md for full schema.

Key fields: phase, taskIndex, totalTasks, taskIteration, maxTaskIterations, awaitingApproval, quickMode, quickAuthorization, phaseSkillLoad, phaseInterview, and optional activePrototypes.

Prototype is an overlay. phase remains one of the five main workflow phases and is never prototype. Resolve basePath before access, treat a missing activePrototypes map as empty, and route every mutation or deletion through hooks/scripts/locked-state.py. Read references/state-file-schema.md before changing state, resuming an overlay, or deciding whether execution state may be deleted.

Commit Behavior

When commitSpec is true:

  1. Stage spec/feature files after generation
  2. Commit with message: chore(<plugin>): commit spec files before implementation
  3. Keep the commit local unless a separate command and explicit authority permit a remote action

When commitSpec is false:

  • Files remain uncommitted
  • User can manually commit later

Prototype terminal records are spec artifacts, so commitSpec controls their local commit. Retained prototype source always receives a local commit on its isolated branch, independent of commitSpec. Neither path authorizes a push, remote branch, PR inclusion, issue mutation, or ticket comment.

Prototype Overlay

The coordinator follows this sequence:

  1. Resolve basePath, reconcile candidates and finals, and reserve or resume activePrototypes.<id> under the state lock. Completion criterion: the main phase and unrelated fields are unchanged.
  2. Keep source in an isolated sibling worktree or eligible scratch path. Completion criterion: the current checkout remains on its original branch and prototype files exist only in the recorded isolation path.
  3. Build and review one falsifiable question. Completion criterion: the reviewer passes the exact candidate bytes, source/run evidence, isolation, blocker, handoff, and sourceDisposition.
  4. Publish one immutable terminal record without overwrite, apply gate selection, restore the recorded return phase/task, then remove the active entry under lock. Completion criterion: downstream work sees only valid, non-superseded, gate-approved evidence.

Normal mode waits without a verdict or handoff timeout and keeps decisions with the user. Quick mode owns the one post-requirements request, can take over the oldest design blocker, uses reviewed receipts for eligible ephemeral cleanup, and always continues to design. Cancellation and interruption preserve source and publish reviewed terminal evidence before active-state removal.

Prototype source, records, cleanup receipts, and quarantines are local evidence. A remote push, branch publication, PR inclusion, issue write, or deletion outside the single quick ephemeral cleanup path requires separate explicit authority.

Task Execution Loop

Ralph Specum v3.0.0+ has a self-contained execution loop via the stop-hook. No external dependencies required.

Key signals:

  • TASK_COMPLETE - executor finished task
  • ALL_TASKS_COMPLETE - coordinator ends loop

Error Handling

When taskIteration > maxTaskIterations: block task, suggest manual intervention.

If state file missing/invalid: output error, suggest re-running implement command.

Branch Management

All Ralph plugins follow consistent branch strategy:

  1. Check current branch before starting
  2. If on default branch (main/master): prompt for branch strategy
  3. If on feature branch: offer to continue or create new
  4. Exact --quick: create a feature branch without prompting only when currently on the default branch; otherwise keep the current non-default branch. Legacy quickMode: true does not select this path.

Prototype isolation is separate from this feature-branch choice. Create or reuse a sibling worktree or eligible scratch path without checking out its branch in the current checkout. Retained and interrupted source remains available for recovery.

Coordinator Behavior

The main agent is a coordinator, not an implementer. Delegate all work to subagents.

Coordinator Responsibilities

  1. Parse user input and determine intent
  2. Read state files for context
  3. Delegate work to subagents via Task tool
  4. Report results to user

Do Not

  • Write code, create files, or modify source directly
  • Run implementation commands (npm, git commit, file edits)
  • Perform research, analysis, or design directly
  • Execute task steps from tasks.md

Delegation Mapping

Work TypeDelegate To
ResearchResearch Team (parallel teammates)
Requirementsproduct-manager subagent
Designarchitect-reviewer subagent
Task planningtask-planner subagent
Task executionspec-executor subagent
Prototype buildingprototype-builder background subagent

Quick mode still requires delegation.

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