agentsclimarketplace

Growth orchestrator

Skill LuckyOneTwoThree/pm-skill/pm-06-growth/orchestrators/growth-orchestrator

当需要制定增长策略或系统化推进增长时使用。增长策略总指挥官,先诊断增长模式,再按需调度获客/激活/留存/变现子编排器。关键词:增长策略、增长模式、AARRR、增长飞轮、增长体系、用户增长、增长瓶颈、增长诊断。From its SKILL.md

Install
npx -y skills add LuckyOneTwoThree/pm-skill --skill growth-orchestrator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 6 stars6 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

12.3 KB, ~4.1k tokens by cl100k_base, as published. Nobody here has run it

增长策略总指挥官

核心原则

先诊断模式,再分发执行

增长不是盲目堆渠道,而是先搞清楚产品适合哪种增长模式,再把资源精准投入到最高杠杆的环节。增长模式决定了获客、激活、留存、变现的策略组合。

编排理念

  1. 模式先行:先诊断增长模式(PLG/SLG/MLG/混合),再决定各环节策略
  2. 杠杆优先:基于飞轮模型识别当前最高杠杆环节,集中资源突破
  3. 数据驱动归因:从增长模式到各环节全链路归因,量化每个动作的贡献
  4. 闭环迭代:增长策略通过数据持续验证和迭代

编排协议

编排协议遵循 orchestrator-protocol.md 统一标准。

Pipeline

pipeline: growth-orchestrator
version: 8.0

post_pipeline:
  - action: stage-summary
    output: output/phase-reports/pm-growth/growth-orchestrator.md

stages:
  - id: phase-1
    name: "增长模式诊断"
    depends_on: []
    skills: [growth-model]
    gate:
      condition: "增长模式已确定,飞轮模型已构建"
      fail_action: "补充产品特征和用户数据"

  - id: phase-2
    name: "获客优化"
    depends_on: [phase-1]
    skills: [acquisition-orchestrator]
    trigger: 获客为瓶颈
    gate:
      condition: "渠道评估完成且漏斗优化方案已生成"
      fail_action: "补充缺失渠道数据或延长分析周期"

  - id: phase-3
    name: "激活优化"
    depends_on: [phase-1]
    skills: [activation-orchestrator]
    trigger: 激活为瓶颈
    gate:
      condition: "Aha Moment候选已识别且Onboarding策略已生成"
      fail_action: "扩大行为搜索范围或补充分群数据"

  - id: phase-4
    name: "留存优化"
    depends_on: [phase-1]
    skills: [retention-orchestrator]
    trigger: 留存为瓶颈
    gate:
      condition: "流失预警模型已构建且用户分层已完成"
      fail_action: "优化模型或补充训练数据"

  - id: phase-5
    name: "变现优化"
    depends_on: [phase-1]
    skills: [revenue-orchestrator]
    trigger: 变现为瓶颈
    gate:
      condition: "付费漏斗分析完成且NRR追踪已建立"
      fail_action: "补充漏斗步骤定义或数据"

  - id: phase-6
    name: "增长策略报告"
    depends_on: [phase-1, phase-2, phase-3, phase-4, phase-5]
    skills: [growth-strategy-report]
    gate:
      condition: "增长策略报告经人类确认"
      fail_action: "调整策略方向和执行路线图"

  - id: phase-7
    name: "GTM策略"
    depends_on: [phase-1]
    skills: [gtm-strategy]
    trigger: 新产品上市/市场拓展
    gate:
      condition: "GTM策略经人类确认"
      fail_action: "确认上市路径和渠道策略"

  - id: phase-8
    name: "运营手册"
    depends_on: [phase-1]
    skills: [product-operations-manual]
    trigger: 运营手册制定需求
    gate:
      condition: "运营手册经人类确认"
      fail_action: "确认运营SOP和应急流程"

阶段执行计划

阶段1:增长模式诊断

调用 growth-model

Skill: growth-model
输入:
  product_features: 用户提供(产品特征)
  user_data: analysis-retention → retention_analysis.yaml
  business_model: 用户提供(商业模式)
输出: output/pm-growth/growth-model/
验证: 北极星指标与≥1个OKR Objective直接关联;增长模型包含≥3个可量化变量;增长飞轮包含≥4个节点且形成闭环;瓶颈约束识别≤5个,每个有量化影响评估
模式: 🤖→👤

阶段2:瓶颈环节优化(条件分支)

根据阶段1诊断的瓶颈环节,调度对应的子编排器。

调用 acquisition-orchestrator

Skill: acquisition-orchestrator
输入:
  growth_model: output/pm-growth/growth-model/
  channel_data: 用户提供
  funnel_data: 用户提供
输出: output/pm-growth/acquisition-analysis/
验证: 渠道评估覆盖19种渠道;获客漏斗各层转化分析完成,优化建议已输出
模式: 🤖→👤

调用 activation-orchestrator

Skill: activation-orchestrator
输入:
  growth_model: output/pm-growth/growth-model/
  user_behavior_data: 用户提供
  retention_data: analysis-retention → retention_analysis.yaml
输出: output/pm-growth/activation-aha/、output/pm-growth/activation-onboarding/
验证: Aha Moment候选已识别;Onboarding策略已生成
模式: 🤖→👤

调用 retention-orchestrator

Skill: retention-orchestrator
输入:
  growth_model: output/pm-growth/growth-model/
  user_behavior_data: 用户提供
  churn_history: 用户提供(流失历史数据)
输出: output/pm-growth/retention-management/
验证: 流失预警模型已构建;用户分层已完成
模式: 🤖→👤

调用 revenue-orchestrator

Skill: revenue-orchestrator
输入:
  growth_model: output/pm-growth/growth-model/
  payment_funnel_data: 用户提供(付费漏斗数据)
  revenue_data: 用户提供(收入数据)
输出: output/pm-growth/revenue-funnel/、output/pm-growth/revenue-nrr/、output/pm-growth/revenue-upsell/
验证: 付费漏斗分析完成;NRR追踪已建立
模式: 🤖→👤

多瓶颈场景:若多个环节均为瓶颈,按飞轮顺序依次调度子编排器(获客→激活→留存→变现)。

阶段3:增长策略报告

调用 growth-strategy-report

Skill: growth-strategy-report
输入:
  growth_model: output/pm-growth/growth-model/
  acquisition_plan: output/pm-growth/acquisition-analysis/(可选)
  activation_plan: output/pm-growth/activation-aha/(可选)
  retention_plan: output/pm-growth/retention-management/(可选)
  revenue_plan: output/pm-growth/revenue-funnel/(可选)
  business_goal: 用户提供(可选)
输出: output/pm-growth/growth-strategy-report/
验证: 飞轮模型完整性(至少3个节点+2条因果关系);策略与瓶颈一致;路线图可执行;漏斗数据完整(AARRR至少3个环节有数据)
模式: 🤖→👤

附加阶段(按需触发)

调用 gtm-strategy

Skill: gtm-strategy
输入:
  positioning: positioning-strategy(可选)
  business_model: business-model-canvas(可选)
  pricing: business-pricing(可选)
  growth_model: output/pm-growth/growth-model/
  product_info: 用户提供
输出: output/pm-growth/gtm-strategy/
验证: ICP画像具体(至少包含行业、规模、角色3个维度);上市路径有依据;渠道预算可执行;成功指标可量化
模式: 🤖→👤

调用 product-operations-manual

Skill: product-operations-manual
输入:
  growth_model: output/pm-growth/growth-model/(可选)
  activation_strategy: output/pm-growth/activation-onboarding/(可选)
  retention_strategy: output/pm-growth/retention-management/(可选)
  revenue_strategy: output/pm-growth/revenue-funnel/(可选)
  product_info: 用户提供
输出: output/pm-growth/product-operations-manual/
验证: SOP可执行;分层策略完整(至少覆盖新/活跃/沉默/流失4类用户);应急流程可操作(P0-P3均有响应SLA和升级路径);模板可直接使用
模式: 🤖→👤

阶段总结(post_pipeline)

所有业务阶段执行完成后,必须立即生成阶段总结文档:

动作: 生成阶段总结
输入:
  所有子Skill输出: output/pm-growth/
  人类决策记录: 本轮执行中的人类决策点及结果
输出: output/phase-reports/pm-growth/growth-orchestrator.md
验证: 阶段总结文档已生成,6项结构(执行概览/关键发现/决策记录/产出清单/风险与待办/下游衔接)均非空
下游衔接:
  primary:
    target: acquisition-orchestrator
    reason: 增长策略制定完成,进入获客优化执行
    input_mapping:
      growth_output: "output/pm-growth/growth-strategy-report/ → acquisition-analysis输入"
  alternatives:
    - target: experiment-orchestrator
      reason: 增长方案需A/B测试验证效果
      condition: 增长方案涉及重大策略变更需量化验证时
    - target: release-orchestrator
      reason: 增长方案已验证,直接全量发布
      condition: 增长方案已有充分数据支撑,无需实验验证时
    - target: growth-orchestrator
      reason: 新产品需上市,进入GTM策略阶段(内部phase-7)
      condition: 增长诊断结论为新产品需上市时
  special_cases: []
模式: 🤖

⏸ 阶段卡口:阶段总结文档已生成且6项结构均非空 → 未通过:补充缺失结构项后重新生成

阶段卡口

卡口条件未通过处理
增长模式诊断完成growth-model输出文件已生成且非空补充产品特征和用户数据
瓶颈环节已识别growth-bottleneck输出文件已生成且非空延长分析周期或扩大数据范围
增长策略报告已确认增长策略报告经人类确认调整策略方向和执行路线图
阶段总结已生成output/phase-reports/pm-growth/growth-orchestrator.md 已生成且6项结构均非空补充缺失结构项后重新生成

人类决策点

决策点触发条件决策内容
增长模式确认growth-model诊断完成确认最终增长模式(PLG/SLG/MLG/混合)
瓶颈优先级确认瓶颈环节识别完成确认资源分配优先级
飞轮模型确认飞轮模型构建完成确认飞轮节点和因果关系
增长策略报告确认growth-strategy-report生成完成确认策略方向和执行路线图
GTM策略确认gtm-strategy生成完成确认上市路径和渠道策略
运营手册确认product-operations-manual生成完成确认运营SOP和应急流程

异常处理

异常类型处理策略
增长模式诊断无法收敛(多种模式得分接近)标注为混合模式,列出各模式得分和依据,由人类决策确认
子编排器执行超时或失败跳过该瓶颈环节,继续执行其他瓶颈环节,最终报告中标注"该环节待补充"
多瓶颈场景下上下文溢出按飞轮顺序仅执行最高优先级瓶颈,其余瓶颈记录待办,分批执行
子Skill输出校验未通过回退至当前阶段重新执行,最多重试1次;仍失败则标记异常并上报人类
上下游数据格式不兼容按下游子Skill输入Schema做字段映射和默认值填充,记录映射关系
阶段总结生成失败基于已完成的子Skill输出生成部分总结,缺失项标注"数据缺失",不阻塞编排完成

变更记录

  • v1.0: 初始版本
  • v2.0: 结构优化
  • v3.0: 新增 growth-strategy-report(增长策略报告)、gtm-strategy(Go-to-Market策略)、product-operations-manual(产品运营手册)
  • v4.0: 编排器优化——任务调度改为阶段执行计划,新增子Skill执行协议,调度规则改为执行模式,阶段卡口和人类决策点改为表格,条件分支子编排器说明
  • v5.0: 执行步骤替换为编排理念,新增异常处理表
  • v6.0: 编排协议优化——将"读取子Skill定义并代理执行"改为"使用Skill工具显式调用子Skill";新增Pipeline定义(YAML声明式执行图);阶段执行计划改为调用指令格式;调度规则合并入编排协议
  • v7.1: 阶段总结强化——Pipeline新增post_pipeline定义;调用规则第6条改为强制执行;阶段执行计划新增阶段总结执行指令;阶段卡口新增阶段总结校验;异常处理新增阶段总结生成失败策略
  • v8.0: 更新引用 acquisition-channel/acquisition-optimize → acquisition-analysis;更新引用 retention-churn/retention-engagement → retention-management;更新输出路径

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.