Back to Details

setup

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

OrchestKit Setup Wizard

OrchestKit 设置向导

Personalized onboarding that scans your codebase, detects your stack, recommends skills and MCPs, and generates an actionable improvement plan.
个性化入门流程，可扫描代码库、检测技术栈、推荐技能与MCP，并生成可执行的改进计划。

When to Use

使用场景

  • First time after installing OrchestKit (
    /plugin install ork
    )
  • Joining a new project and want OrchestKit tuned for it
  • Periodically with 
    --rescan
    to track improvement
  • Enterprise users who need safe, user-scoped install confirmation
  • 安装OrchestKit后首次使用（
    /plugin install ork
    ）
  • 加入新项目时，需要为其适配OrchestKit
  • 定期使用
    --rescan
    参数跟踪改进情况
  • 企业用户需要安全的、用户范围的安装确认

Quick Start

快速开始

bash
/ork:setup              # Full 8-phase wizard
/ork:setup --rescan     # Re-scan after changes (skip safety phase)
/ork:setup --score-only # Just show readiness score
/ork:setup --plan-only  # Just show improvement plan
/ork:setup --channel    # Just show release channel
/ork:setup --configure  # Jump directly to Phase 3.5: project configuration wizard
bash
/ork:setup              # 完整的8阶段向导
/ork:setup --rescan     # 变更后重新扫描（跳过安全阶段）
/ork:setup --score-only # 仅显示就绪度评分
/ork:setup --plan-only  # 仅显示改进计划
/ork:setup --channel    # 仅显示发布渠道
/ork:setup --configure  # 直接跳转到3.5阶段：项目配置向导

Argument Resolution

参数解析

python
FLAG = "$ARGUMENTS[0]"  # First token: --rescan, --score-only, --plan-only, --channel
python
FLAG = "$ARGUMENTS[0]"  # 第一个参数：--rescan, --score-only, --plan-only, --channel

If no arguments, run full 8-phase wizard.

若无参数，运行完整的8阶段向导。

$ARGUMENTS is the full string (CC 2.1.59 indexed access)

$ARGUMENTS为完整参数字符串（CC 2.1.59 索引访问）

undefined
undefined

The Nine Phases

十个阶段

PhaseWhatTools UsedOutput
1. ScanDetect languages, frameworks, infra, existing configGlob, Grep, ReadRaw scan data
2. StackClassify detected stack, confidence levels—Stack profile
3. SafetyCheck existing config, confirm scope (user/project)Read, AskUserQuestionInstall confirmation
3.5. ConfigureInteractive project configuration wizard → writes env block to per-project settingsRead, Write, AskUserQuestionConfigured settings file
4. SkillsMatch stack to skills, suggest custom skillsGrep, GlobSkill recommendations
5. MCPsRecommend MCPs based on stack and gapsRead, BashMCP recommendations
6. ScoreCompute readiness score (0-10, 6 dimensions)All above dataReadiness score
7. PlanPrioritized improvements with runnable commands—Action plan
8. KeysInstall recommended keyboard shortcutsRead, Bash, AskUserQuestionKeybindings
阶段内容使用工具输出结果
1. 扫描检测语言、框架、基础设施及现有配置Glob, Grep, Read原始扫描数据
2. 技术栈对检测到的技术栈进行分类，给出置信度—技术栈配置文件
3. 安全检查检查现有配置，确认作用范围（用户/项目）Read, AskUserQuestion安装确认信息
3.5. 配置交互式项目配置向导 → 将环境块写入项目专属设置Read, Write, AskUserQuestion已配置的设置文件
4. 技能推荐匹配技术栈与技能，建议自定义技能Grep, Glob技能推荐列表
5. MCP推荐根据技术栈与缺口推荐MCPRead, BashMCP推荐列表
6. 就绪度评分计算就绪度评分（0-10分，6个维度）以上所有数据就绪度评分
7. 改进计划生成带可执行命令的优先级改进方案—行动计划
8. 快捷键配置安装推荐的键盘快捷键Read, Bash, AskUserQuestion快捷键绑定
9. 遥测与Webhook配置遥测事件与Webhook集成Read, AskUserQuestion, Bash遥测配置
10. 可选集成配置Agentation UI、多仓库工作区等高级功能Read, Bash集成配置

Phase 1: Scan

阶段1：扫描

Load details: 
Read("${CLAUDE_SKILL_DIR}/references/scan-phase.md")
for full scan commands (20 parallel Glob probes + dependency file reads + pattern detection counts).
Scans for package manifests (package.json, pyproject.toml, go.mod, Cargo.toml, etc.), infrastructure (Docker, GitHub Actions, Terraform, K8s), and existing CC configuration. Pattern detection counts API routes, React components, models, and tests for custom skill suggestions.
加载详细说明：
Read("${CLAUDE_SKILL_DIR}/references/scan-phase.md")
获取完整扫描命令（20个并行Glob探测 + 依赖文件读取 + 模式检测统计）。
扫描包清单（package.json、pyproject.toml、go.mod、Cargo.toml等）、基础设施（Docker、GitHub Actions、Terraform、K8s）及现有CC配置。通过模式检测统计API路由、React组件、模型和测试数量，为自定义技能提供建议。

Phase 2: Stack Detection

阶段2：技术栈检测

Classify scan results into a stack profile and present to user (languages, frameworks, database, infra, testing, existing CC config).
Load 
Read("${CLAUDE_SKILL_DIR}/references/stack-skill-mapping.md")
for the full stack-to-skill mapping table, MCP recommendation matrix, and custom skill suggestion patterns.
将扫描结果分类为技术栈配置文件并展示给用户（语言、框架、数据库、基础设施、测试、现有CC配置）。
加载
Read("${CLAUDE_SKILL_DIR}/references/stack-skill-mapping.md")
获取完整的技术栈-技能映射表、MCP推荐矩阵及自定义技能建议模式。

Phase 2b: Channel Detection

阶段2b：渠道检测

Detect release channel from 
manifests/ork.json
version string. Classification: 
X.Y.Z
= stable, 
X.Y.Z-beta.*
= beta, 
X.Y.Z-alpha.*
= alpha. Display alongside stack profile. Use 
--channel
flag to show only channel detection.
从
manifests/ork.json
版本字符串中检测发布渠道。分类规则：
X.Y.Z
= 稳定版，
X.Y.Z-beta.*
= 测试版，
X.Y.Z-alpha.*
= 预览版。与技术栈配置文件一同展示。使用
--channel
参数可仅显示渠道检测结果。

Phase 3: Safety Check

阶段3：安全检查

Load details: 
Read("${CLAUDE_SKILL_DIR}/references/safety-check.md")
for the full AskUserQuestion prompt and conflict detection logic.
Offers three install scopes: User-only (recommended, invisible to teammates), Project-wide (committed to repo), or Already installed (skip to configure). Checks for existing OrchestKit installs and conflicting plugins.
Console auth (CC 2.1.79+): Users authenticating via Anthropic Console (API billing) can run 
claude auth login --console
instead of the default OAuth flow. Useful for teams on usage-based API billing rather than Pro/Team subscriptions.
加载详细说明：
Read("${CLAUDE_SKILL_DIR}/references/safety-check.md")
获取完整的AskUserQuestion提示语与冲突检测逻辑。
提供三种安装范围：仅用户（推荐，对团队成员不可见）、项目全局（提交到仓库）、已安装（直接跳转到配置阶段）。检查现有OrchestKit安装及冲突插件。
控制台认证（CC 2.1.79+）： 通过Anthropic Console（API计费）认证的用户可运行
claude auth login --console
替代默认OAuth流程。适用于采用基于使用量计费而非Pro/Team订阅的团队。

Phase 3.5: Project Configuration Wizard

阶段3.5：项目配置向导

Load details: 
Read("${CLAUDE_SKILL_DIR}/references/configure-wizard.md")
for the full 6-step interactive configuration flow (branch strategy, commit scope, localhost browser, perf telemetry, log verbosity, webhook telemetry) and env var reference.
Also reachable directly via 
/ork:setup --configure
— skips phases 1-3.
加载详细说明：
Read("${CLAUDE_SKILL_DIR}/references/configure-wizard.md")
获取完整的6步交互式配置流程（分支策略、提交范围、本地浏览器、性能遥测、日志 verbosity、Webhook遥测）及环境变量参考。
也可通过
/ork:setup --configure
直接进入 — 跳过第1-3阶段。

Phase 4: Skill Recommendations

阶段4：技能推荐

Present skill categories using 
AskUserQuestion
with 4 focus options (Full-stack, Backend, Frontend, DevOps) with 
multiSelect: true
. Load 
Read("${CLAUDE_SKILL_DIR}/references/stack-skill-mapping.md")
for mapping tables and custom skill suggestions.
使用AskUserQuestion展示技能分类，提供4个聚焦选项（全栈、后端、前端、DevOps），支持
multiSelect: true
。加载
Read("${CLAUDE_SKILL_DIR}/references/stack-skill-mapping.md")
获取映射表及自定义技能建议。

Phase 5: MCP Recommendations

阶段5：MCP推荐

Check installed vs recommended MCPs by reading 
.mcp.json
and 
~/.claude/settings.json
. Load 
Read("${CLAUDE_SKILL_DIR}/references/stack-skill-mapping.md")
for the MCP recommendation matrix. Present as toggles with install commands.
通过读取
.mcp.json
和
~/.claude/settings.json
检查已安装与推荐的MCP。加载
Read("${CLAUDE_SKILL_DIR}/references/stack-skill-mapping.md")
获取MCP推荐矩阵。以开关形式展示并提供安装命令。

Phase 6: Readiness Score

阶段6：就绪度评分

Compute a composite score (0-10) from 6 dimensions. Load 
Read("${CLAUDE_SKILL_DIR}/references/readiness-scoring.md")
for dimension weights, score presentation template, memory integration, and improvement plan template.
从6个维度计算综合评分（0-10分）。加载
Read("${CLAUDE_SKILL_DIR}/references/readiness-scoring.md")
获取维度权重、评分展示模板、内存集成及改进计划模板。

Phase 7: Improvement Plan

阶段7：改进计划

Generate prioritized, runnable recommendations in P0/P1/P2 tiers. See 
readiness-scoring.md
for the template and memory persistence pattern.
生成按优先级划分的可执行建议，分为P0/P1/P2等级。详见
readiness-scoring.md
中的模板与内存持久化模式。

Phase 7b: CLAUDE.md Health Check

阶段7b：CLAUDE.md健康检查

After the improvement plan, check if the user's CLAUDE.md could benefit from CC 2.1.59+ modular structure.
Load details: 
Read("${CLAUDE_SKILL_DIR}/references/claude-md-health.md")
for analysis steps, thresholds, @import syntax, and 
.claude/rules/
path-scoped rules.
python
undefined
在改进计划之后，检查用户的CLAUDE.md是否可受益于CC 2.1.59+的模块化结构。
加载详细说明：
Read("${CLAUDE_SKILL_DIR}/references/claude-md-health.md")
获取分析步骤、阈值、@import语法及
.claude/rules/
路径作用域规则。
python
undefined

Quick check

快速检查

Bash(command="wc -l CLAUDE.md 2>/dev/null | awk '{print $1}'") Glob(pattern=".claude/rules/*.md")

If CLAUDE.md > 200 lines and no `.claude/rules/` exist, recommend splitting. Show the output template from the reference doc.
Bash(command="wc -l CLAUDE.md 2>/dev/null | awk '{print $1}'") Glob(pattern=".claude/rules/*.md")

若CLAUDE.md超过200行且不存在`.claude/rules/`目录，建议拆分文件。展示参考文档中的输出模板。

Phase 8: Keybindings

阶段8：快捷键配置

Load details: 
Read("${CLAUDE_SKILL_DIR}/references/keybindings.md")
for the full keybinding prompt, default shortcuts, and merge logic.
Offers 5 recommended shortcuts (commit, verify, implement, explore, review-pr). Merges with existing 
~/.claude/keybindings.json
without overwriting user-defined bindings.
加载详细说明：
Read("${CLAUDE_SKILL_DIR}/references/keybindings.md")
获取完整的快捷键提示语、默认快捷键及合并逻辑。
提供5个推荐快捷键（提交、验证、实现、探索、PR评审）。与现有
~/.claude/keybindings.json
合并，不会覆盖用户自定义绑定。

Post-Setup

设置后提示

Tip (CC 2.1.69+): After setup completes, run 
/reload-plugins
to activate all plugin changes without restarting your session.
提示（CC 2.1.69+）： 设置完成后，运行
/reload-plugins
可激活所有插件变更，无需重启会话。

Phase 9: Telemetry & Webhooks

阶段9：遥测与Webhook

Previously in 
/ork:configure
. Now part of setup for single entry point.
Load details: 
Read("${CLAUDE_SKILL_DIR}/references/telemetry-setup.md")
for the full configuration flow.
Ask user preference with AskUserQuestion:
ModeEventsAuthOverhead
Full streamingAll 18 CC events via HTTP hooksBearer tokenNear-zero
Summary onlySessionEnd + worktree eventsHMAC authNone
SkipNo telemetry—None
If streaming selected:
  1. Ask for webhook URL
  2. Run 
    npm run generate:http-hooks -- <url> --write
  3. Save to 
    .claude/orchestration/config.json
  4. Remind about 
    ORCHESTKIT_HOOK_TOKEN
    env var
原属于
/ork:configure
。现在整合到设置流程中，提供统一入口。
加载详细说明：
Read("${CLAUDE_SKILL_DIR}/references/telemetry-setup.md")
获取完整配置流程。
通过AskUserQuestion询问用户偏好：
模式事件认证开销
全量流所有18种CC事件通过HTTP钩子传输Bearer令牌近乎为零
仅摘要SessionEnd + 工作区事件HMAC认证无
跳过无遥测—无
若选择全量流：
  1. 询问Webhook URL
  2. 运行
    npm run generate:http-hooks -- <url> --write
  3. 保存到
    .claude/orchestration/config.json
  4. 提醒设置
    ORCHESTKIT_HOOK_TOKEN
    环境变量

Phase 10: Optional Integrations

阶段10：可选集成

Load details: 
Read("${CLAUDE_SKILL_DIR}/references/integrations.md")
for setup steps.
Covers Agentation UI annotation tool, CC version-specific settings (CC 2.1.7 MCP deferral, CC 2.1.20 task deletion, CC 2.1.23 spinner verbs, CC 2.1.79 turn duration display), and monorepo worktree optimization.
加载详细说明：
Read("${CLAUDE_SKILL_DIR}/references/integrations.md")
获取设置步骤。
涵盖Agentation UI标注工具、CC版本专属设置（CC 2.1.7 MCP延迟、CC 2.1.20任务删除、CC 2.1.23 spinner verbs、CC 2.1.79对话时长显示）及多仓库工作区优化。

Monorepo Sparse Paths (CC 2.1.76+)

多仓库稀疏路径（CC 2.1.76+）

If Phase 1 scan detected a monorepo (pnpm-workspace.yaml, nx.json, lerna.json, turbo.json, or package.json workspaces), suggest configuring 
worktree.sparsePaths
in 
.claude/settings.json
:
json
{
  "worktree": {
    "sparsePaths": ["src/", "packages/core/", "tests/", "scripts/"]
  }
}
This makes 
--worktree
and agent 
isolation: worktree
check out only the listed directories via git sparse-checkout — significantly faster in large monorepos.
若阶段1扫描检测到多仓库（pnpm-workspace.yaml、nx.json、lerna.json、turbo.json或package.json workspaces），建议在
.claude/settings.json
中配置
worktree.sparsePaths
：
json
{
  "worktree": {
    "sparsePaths": ["src/", "packages/core/", "tests/", "scripts/"]
  }
}
此配置可让
--worktree
和agent的
isolation: worktree
通过git sparse-checkout仅检出指定目录 — 在大型多仓库中可显著提升速度。

Multi-Directory Plugin Seeding (CC 2.1.79+)

多目录插件种子（CC 2.1.79+）

CLAUDE_CODE_PLUGIN_SEED_DIR
now supports multiple directories separated by the platform path delimiter (
:
on Unix, 
;
on Windows):
bash
export CLAUDE_CODE_PLUGIN_SEED_DIR="/shared/team/plugins:/shared/org/plugins"
This is useful for teams with multiple plugin sources — e.g., a shared org-wide plugin directory alongside a team-specific one. Claude Code scans all listed directories for plugin definitions at startup.
CLAUDE_CODE_PLUGIN_SEED_DIR
现在支持通过平台路径分隔符（Unix为
:
，Windows为
;
）分隔多个目录：
bash
export CLAUDE_CODE_PLUGIN_SEED_DIR="/shared/team/plugins:/shared/org/plugins"
这对拥有多个插件源的团队非常有用 — 例如，将组织级共享插件目录与团队专属插件目录结合使用。Claude Code启动时会扫描所有列出的目录以查找插件定义。

CLI Flags

CLI参数

FlagBehavior
(none)Full wizard (phases 1-10)
--rescan
Re-run scan + score, skip safety phase
--configure
Jump directly to Phase 3.5: project configuration wizard
--score-only
Show current readiness score (Phase 6 only)
--plan-only
Show improvement plan (Phase 7 only)
--channel
Show detected release channel only
--telemetry
Jump to Phase 9: telemetry/webhook setup
--preset
Apply a preset (complete/standard/lite/hooks-only/monorepo)
参数行为
(无)完整向导（第1-10阶段）
--rescan
重新运行扫描+评分，跳过安全阶段
--configure
直接跳转到3.5阶段：项目配置向导
--score-only
仅显示当前就绪度评分（仅第6阶段）
--plan-only
仅显示改进计划（仅第7阶段）
--channel
仅显示检测到的发布渠道
--telemetry
跳转到第9阶段：遥测/Webhook设置
--preset
应用预设配置（complete/standard/lite/hooks-only/monorepo）

Presets (via --preset)

预设配置（通过--preset）

Apply a preset to quickly configure OrchestKit without the full wizard:
PresetSkillsAgentsHooksBest For
complete913196Full power — everything enabled
standard91096Skills + hooks, no agents
lite10096Essential workflow skills only
hooks-only0096Just safety hooks
monorepo913196Complete + monorepo workspace detection
Load preset details: 
Read("${CLAUDE_SKILL_DIR}/references/presets.md")
应用预设可快速配置OrchestKit，无需完整向导：
预设技能数量Agent数量Hooks数量适用场景
complete913196全功能 — 启用所有选项
standard91096技能+Hooks，无Agent
lite10096仅核心工作流技能
hooks-only0096仅安全Hooks
monorepo913196完整配置+多仓库工作区检测
加载预设详情：
Read("${CLAUDE_SKILL_DIR}/references/presets.md")

References

参考文档

Load on demand with 
Read("${CLAUDE_SKILL_DIR}/references/<file>")
:
FileContent
scan-phase.md
Phase 1: 20 parallel Glob probes + pattern detection
safety-check.md
Phase 3: Install scope and conflict detection
configure-wizard.md
Phase 3.5: 6-step interactive project config
claude-md-health.md
Phase 7b: CLAUDE.md modular structure analysis
keybindings.md
Phase 8: Keyboard shortcut recommendations
telemetry-setup.md
Phase 9: Webhook/telemetry configuration
integrations.md
Phase 10: Agentation + CC version settings
presets.md
Preset definitions and skill/agent matrices
通过
Read("${CLAUDE_SKILL_DIR}/references/<file>")
按需加载：
文件内容
scan-phase.md
阶段1：20个并行Glob探测+模式检测
safety-check.md
阶段3：安装范围与冲突检测
configure-wizard.md
阶段3.5：6步交互式项目配置
claude-md-health.md
阶段7b：CLAUDE.md模块化结构分析
keybindings.md
阶段8：键盘快捷键推荐
telemetry-setup.md
阶段9：Webhook/遥测配置
integrations.md
阶段10：Agentation+CC版本设置
presets.md
预设定义与技能/Agent矩阵

Related Skills

相关技能

  • ork:doctor
    — Health diagnostics (wizard uses its checks)
  • ork:configure
    — Internal configuration (called by wizard phases 3.5, 9, 10)
  • ork:remember
    — Knowledge persistence (wizard seeds initial patterns)
  • ork:explore
    — Deep codebase analysis (wizard links for follow-up)
  • ork:help
    — Skill directory (wizard surfaces relevant subset)
  • ork:doctor
    — 健康诊断（向导会使用其检查逻辑）
  • ork:configure
    — 内部配置（被向导的3.5、9、10阶段调用）
  • ork:remember
    — 知识持久化（向导会初始化模式）
  • ork:explore
    — 深度代码库分析（向导提供后续链接）
  • ork:help
    — 技能目录（向导会展示相关子集）