一键导入
sh-help
List all 34 Shepherd commands organized by category
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
List all 34 Shepherd commands organized by category
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Multi-expert AI/ML specification review with scoring gate — model architecture, evaluation rigor, safety/alignment, and production readiness for AI systems and LLM apps
Multi-expert architecture review — boundaries, integration patterns, failure modes, evolvability
Multi-expert business analysis with advisory recommendations (no scoring gate)
Multi-expert intelligence pipeline review — discovery quality, ingestion resilience, scoring validity, platform compliance, taxonomy coherence, cost efficiency
Multi-expert mobile/native app specification review with scoring gate — Android, iOS, Swift, SwiftUI
Multi-expert personal-development review with scoring gate — learnability, adoption, human-centeredness, and capability impact for talent/learning/AI-augmentation designs
| name | sh-help |
| description | List all 34 Shepherd commands organized by category |
Print the following command reference exactly:
| Command | Description |
|---|---|
/sh:triage | Triage INBOX items into BACKLOG with priority and categorization |
/sh:brainstorm | Divergent ideation session — generate and explore possibilities |
/sh:spec-review | Single-reviewer specification quality check with pass/fail |
/sh:workflow | Move BACKLOG items through Ideation > Refining > Ready pipeline |
/sh:kickoff | Restore session context and resume work from handoff state |
| Command | Description |
|---|---|
/sh:autopilot | Autonomous execution loop — read task, execute, commit, repeat |
/sh:loop | Generic iteration loop with configurable exit conditions |
/sh:dor | Definition of Ready — validate task meets entry criteria |
/sh:dod | Definition of Done — validate deliverable meets quality gate |
/sh:retro | Retrospective — reflect on completed work and capture learnings |
/sh:handoff | Save full session context for cross-session persistence |
| Command | Description |
|---|---|
/sh:plan | Create implementation plan with tasks, dependencies, and estimates |
/sh:execute | Execute a planned task with verification and quality checks |
/sh:tdd | Test-driven development — red/green/refactor cycle |
/sh:verify | Run verification suite and report pass/fail status |
/sh:review | Code review with actionable feedback and severity ratings |
superpowers:finishing-a-development-branch | Finalize deliverable — commit, update docs, move to DONE |
/sh:worktree | Manage git worktrees for parallel development branches |
/sh:parallel | Execute independent tasks concurrently across worktrees |
| Command | Description |
|---|---|
/sh:analyze | Deep analysis of code, architecture, or system behavior |
/sh:debug | Systematic debugging with hypothesis testing and root cause analysis |
/sh:troubleshoot | Guided troubleshooting for infrastructure and environment issues |
/sh:research | Research a topic with source evaluation and synthesis |
/sh:explain | Explain code, concept, or architecture with appropriate depth |
/sh:estimate | Estimate effort, complexity, and risk for a task or feature |
/sh:test | Generate or run test suites with coverage analysis |
/sh:document | Generate documentation from code or specifications |
/sh:index-repo | Index repository structure for fast context retrieval |
/sh:select-tool | Recommend the best tool or approach for a given problem |
| Command | Description |
|---|---|
/sh:ux-design | Build clickable wireframe prototypes in code — validate flow before hi-fi |
| Command | Description |
|---|---|
/sh:spec-panel | Multi-expert specification review with scoring gate >= 7.0 |
/sh:business-panel | Multi-expert business analysis with advisory recommendations |
| Command | Description |
|---|---|
/sh:ralph | Start persistent Ralph loop for iterative task completion |
/sh:cancel | Cancel active Ralph loop and report iteration count |
| Command | Description |
|---|---|
/sh:help | List all Shepherd commands (this output) |
Shepherd v2.0 — AgentFlow + Superpowers + SuperClaude + Ralph Loop