Preamble (run first)
# 检查是否有相关前置文档
if [ -f "docs/05-产品战略/产品组合战略.md" ]; then
echo "✅ 检测到产品组合战略文档"
fi
if [ -f "docs/05-产品战略/资源分配方案.md" ]; then
echo "✅ 检测到资源分配方案文档"
fi
Boundaries when writing
When only discussing, do not write files; startup checks do not create directories. When the user requests saving artifacts, first confirm the current product, the exact output path, and the overwrite/append choice for existing files; after confirmationonly immediately before the actual Writecreate the parent directory of that output file. For complete installation, see skills/_shared/artifact-policy.md, installing this skill alone also follows this paragraph.
Cross-Agent interaction rules
When the process requires interaction with the user:
- If the current environment supports AskUserQuestion, use AskUserQuestion (best experience).
- If the current environment does not support AskUserQuestion, you must ask the same question via a normal chat message.
- Ask only one question at a time.
- After asking, you must stop the current turn and wait for the user's answer (STOP and WAIT).
- Do not generate documents or write to docs before the user answers.
- Existing docs files cannot replace the user's answer in this turn.
Execution process
Step 1: Clarify the decision question
How to ask: Use AskUserQuestion to ask one by one
Question 1: Decision type
Use AskUserQuestion to ask:
What type of decision are you facing?
A) Build vs outsource decision B) Build vs acquire decision C) Product investment decision D) Market entry decision E) Business transformation decision F) Partner selection G) Other (please enter manually)
Question 2: Decision background
Ask:
Please briefly describe the decision background
Guide the user to input:
- Decision trigger factors
- Problems or opportunities faced
- Decision time requirements
- Scope of decision impact
Question 3: Decision constraints
Ask:
What are the main constraints the decision faces? (multiple choice)
A) Funding budget constraints B) Time urgency C) Technical capability constraints D) Talent resource constraints E) Risk tolerance F) Strategic alignment requirements G) Other (please enter manually)
Step 2: Identify decision options
Based on the decision type, identify the available options:
Scenario A: Build vs outsource decision
Available options：
- Fully in-house
- Partially outsourced
- Fully outsourced
- Hybrid model (in-house core, outsource non-core)
Scenario B: Build vs acquire decision
Available options：
- Fully in-house
- Acquire a mature company
- Equity investment
- Strategic cooperation
- Patent licensing
Scenario C: Product investment decision
Available options：
- Major investment (go all in)
- Moderate investment (steady progress)
- Small-scale trial (rapid validation)
- No investment for now (wait and see)
Step 3: Decision matrix evaluation
For each decision option, conduct a multi-dimensional evaluation:
Evaluation dimension definition：
Ask:
Please select the decision evaluation dimensions
Recommended dimensions:
- Financial dimension: cost, benefit, ROI
- Time dimension: time to market, development cycle
- Technical dimension: technical risk, technical controllability
- Strategic dimension: strategic alignment, competitive advantage
- Resource dimension: human resource requirements, funding requirements
- Risk dimension: market risk, technical risk, execution risk
Item-by-item evaluation：
For each dimension, use AskUserQuestion to ask for the score of each option (1-10 points):
Example (in-house development vs outsourcing)：
Please score the "fully in-house development" option on the "financial cost" dimension (1-10 points; the lower the score, the higher the cost)
A) 1-2 points (extremely high cost) B) 3-4 points (relatively high cost) C) 5-6 points (moderate cost) D) 7-8 points (relatively low cost) E) 9-10 points (extremely low cost)
Evaluate all options and all dimensions in sequence.
Weight setting：
Ask:
How should the weights of each evaluation dimension be allocated?
Guide the user to allocate weights (total 100%):
| Dimension | Weight |
|---|---|
| Financial cost | [X]% |
| Time to market | [X]% |
| Technical risk | [X]% |
| Strategic alignment | [X]% |
| Long-term competitiveness | [X]% |
| Total | 100% |
Step 4: Decision tree analysis
For complex decisions, build a decision tree:
Decision tree construction：
Ask about key decision nodes:
What are the key uncertainty factors in the decision?
Example：
- Will market demand meet the target? (Probability: [X]%)
- Can the technical challenges be overcome? (Probability: [X]%)
- Will competitors enter the market early? (Probability: [X]%)
Scenario analysis：
For each uncertainty factor, analyze different scenarios:
Scenario 1: Market demand meets the target
- Probability: [X]%
- Benefit: [X] ten thousand yuan
- Cost: [X] ten thousand yuan
- Net benefit: [X] ten thousand yuan
Scenario 2: Market demand does not meet the target
- Probability: [X]%
- Benefit: [X] ten thousand yuan
- Cost: [X] ten thousand yuan
- Net benefit: [X] ten thousand yuan
Expected benefit: probability-weighted average
Step 5: Sensitivity analysis
Test the impact of changes in key assumptions on the decision outcome:
Sensitivity testing：
Ask:
Which key assumptions may affect the decision outcome?
Common key assumptions：
- Development cost: budget [X] ten thousand yuan
- Development time: [X] months
- Market size: [X] hundred million yuan
- Market growth rate: [X]%
- Competitor dynamics: entry within [X] months
Scenario testing：
Scenario 1: Development cost increases by 20%
- Impact on each option: [analysis]
- Whether the decision changes: [yes/no]
Scenario 2: Market size decreases by 30%
- Impact on each option: [analysis]
- Whether the decision changes: [yes/no]
Step 6: Generate strategic decision recommendations
Use the Write tool to generate the document:
---
product: [产品/项目名称]
created_at: [当前时间]
author: [用户]
skill: pm-decision
status: draft
---
# 战略决策建议报告
## 一、决策背景
### 决策问题
**核心问题**：[自研 vs 外包决策]
**决策背景**：
[描述决策背景，包括触发因素、面临的问题或机会]
**决策时间要求**：[X]周内做出决策
**决策影响范围**：[X]个部门，影响[X]人
---
### 决策约束条件
1. **资金预算限制**：预算上限[X]万元
2. **时间紧迫性**：需在[X]个月内上线
3. **技术能力限制**：[描述]
4. **人才资源限制**：现有团队[X]人
5. **风险承受能力**：[低/中/高]
---
## 二、决策选项
### 方案1：完全自研
**方案描述**：
组建[X]人团队，完全自主开发，预计[X]个月完成。
**优势**：
- 技术完全可控
- 知识产权自有
- 可深度定制
**劣势**：
- 成本高（[X]万元）
- 周期长（[X]个月）
- 技术风险高
---
### 方案2：部分外包
**方案描述**：
核心模块自研，非核心模块外包给[供应商名称]，预计[X]个月完成。
**优势**：
- 平衡成本与可控性
- 加快开发速度
- 降低技术风险
**劣势**：
- 沟通协调成本
- 依赖外部供应商
- 知识产权需明确
---
### 方案3：完全外包
**方案描述**：
整体外包给[供应商名称]，预计[X]个月完成，总价[X]万元。
**优势**：
- 成本可控
- 时间最短
- 风险转移
**劣势**：
- 技术不可控
- 后续迭代依赖供应商
- 知识产权风险
---
## 三、决策矩阵评估
### 评估维度与权重
| 维度 | 权重 | 说明 |
|------|------|------|
| 财务成本 | 25% | 开发成本、运维成本 |
| 上市时间 | 20% | 开发周期、上线时间 |
| 技术风险 | 20% | 技术难度、技术可控性 |
| 战略一致性 | 20% | 与公司战略契合度 |
| 长期竞争力 | 15% | 技术积累、竞争壁垒 |
| **合计** | 100% | - |
---
### 评分结果
| 维度 | 权重 | 完全自研 | 部分外包 | 完全外包 |
|------|------|---------|---------|---------|
| 财务成本 | 25% | 4分 | 6分 | 8分 |
| 上市时间 | 20% | 5分 | 7分 | 9分 |
| 技术风险 | 20% | 6分 | 7分 | 5分 |
| 战略一致性 | 20% | 9分 | 7分 | 4分 |
| 长期竞争力 | 15% | 9分 | 7分 | 3分 |
| **加权总分** | - | **6.4分** | **6.75分** | **5.95分** |
---
### 评估结论
**推荐方案**：部分外包（得分最高：6.75分）
**理由**：
1. 在财务成本、时间、风险之间取得平衡
2. 核心技术自主可控，保障战略一致性
3. 非核心模块外包，加快上市速度
---
## 四、决策树分析
### 决策树结构
Decision Node ├─ Option 1: Fully in-house development │ ├─ Technical success (probability 70%) → Gain: [X] ten-thousand yuan │ └─ Technical failure (probability 30%) → Loss: [X] ten-thousand yuan │ └─ Expected gain: [X] ten-thousand yuan │ ├─ Option 2: Partial outsourcing │ ├─ Execution success (probability 80%) → Gain: [X] ten-thousand yuan │ └─ Execution failure (probability 20%) → Loss: [X] ten-thousand yuan │ └─ Expected gain: [X] ten-thousand yuan │ └─ Option 3: Full outsourcing ├─ Quality meets standard (probability 60%) → Gain: [X] ten-thousand yuan └─ Quality does not meet standard (probability 40%) → Loss: [X] ten-thousand yuan └─ Expected gain: [X] ten-thousand yuan
---
### 期望收益对比
| 方案 | 成功概率 | 成功收益 | 失败概率 | 失败损失 | 期望收益 |
|------|---------|---------|---------|---------|---------|
| 完全自研 | 70% | [X]万 | 30% | [X]万 | [X]万 |
| 部分外包 | 80% | [X]万 | 20% | [X]万 | [X]万 |
| 完全外包 | 60% | [X]万 | 40% | [X]万 | [X]万 |
**结论**：部分外包期望收益最高
---
## 五、敏感性分析
### 关键假设
1. 开发成本：预算[X]万元
2. 开发时间：[X]个月
3. 市场规模：[X]亿元
4. 技术难度：中
---
### 情景测试
#### 情景1：开发成本增加20%
**影响分析**：
- 完全自研：成本增加[X]万，得分降低至[X]分
- 部分外包：成本增加[X]万，得分降低至[X]分
- 完全外包：成本固定，得分不变
**决策改变**：否，仍推荐部分外包
---
#### 情景2：市场规模下降30%
**影响分析**：
- 各方案收益下降30%
- ROI降低，但仍为正向
- 决策框架不变
**决策改变**：否，仍推荐部分外包
---
#### 情景3：技术难度升级为"高"
**影响分析**：
- 完全自研风险大幅增加
- 部分外包风险适度增加
- 完全外包风险较低
**决策改变**：可能倾向于完全外包，但需权衡知识产权风险
---
## 六、风险评估
### 风险1: 供应商选择不当
**影响**：项目延期或质量不达标
**概率**：中
**应对**：
1. 严格的供应商筛选流程
2. 分阶段验收机制
3. 备选供应商名单
---
### 风险2: 核心人员流失
**影响**：核心技术无法延续
**概率**：低
**应对**：
1. 关键知识文档化
2. 核心人员激励
3. 团队备份机制
---
### 风险3: 市场需求变化
**影响**：产品不再符合市场需求
**概率**：中
**应对**：
1. 敏捷开发，快速迭代
2. 定期市场调研
3. MVP快速验证
---
## 七、决策建议
### 最终推荐
**推荐方案**：部分外包
**核心理由**：
1. 决策矩阵得分最高（6.75分）
2. 期望收益最高（[X]万元）
3. 在成本、时间、风险、战略之间取得平衡
---
### 实施路径
**阶段1：供应商筛选（Week 1-2）**
- 发布需求文档
- 供应商报价与评估
- 最终选择
**阶段2：合同签订（Week 3）**
- 明确交付标准
- 分阶段付款
- 知识产权归属
**阶段3：开发执行（Week 4-12）**
- 每周进度检查
- 分阶段验收
- 风险监控
**阶段4：验收上线（Week 13-14）**
- 全面测试
- 问题修复
- 正式上线
---
### 成功标准
- 按时交付：[X]周内完成
- 质量达标：无P0级Bug
- 成本可控：不超预算10%
- 知识产权：核心模块知识产权归公司所有
---
## 八、下一步建议
建议执行：
1. /pm-resource - 制定详细的资源分配方案
2. /pm-risk - 制定风险管控方案
3. /pm-release - 制定上线执行方案
---
**项目状态**: 战略决策分析完成
**生成时间**: [当前时间]
**生成工具**: super-pm
Subagent Parallel Analysis (new in v2.0)
After the decision matrix evaluation is completed, subagents can be dispatched to conduct in-depth analysis in parallel:
Agent 1: Option Feasibility Research
type: "general-purpose"
prompt: "基于决策选项，搜索各方案的实际案例、成功率和关键因素。搜索工具优先级：首选 AnySearch（Bash 调用 anysearch_cli.py（先定位：检查 ~/.claude/skills/anysearch/scripts/ ~/.opencode/skills/anysearch/scripts/ ~/.openclaw/skills/anysearch/scripts/ ~/.cursor/skills/anysearch/scripts/ ~/.anysearch/scripts/ 或 which anysearch_cli.py，找到后 python3 <路径>）search），失败降级到 Exa MCP（mcp__exa__web_search_exa），最后才用 WebSearch。使用 WebSearch 时标注降级模式。"
Agent 2: Market Data Validation
type: "general-purpose"
prompt: "搜索行业数据验证决策假设，包括市场规模、竞争格局、增长率。搜索工具优先级：首选 AnySearch（Bash 调用 anysearch_cli.py（先定位：检查 ~/.claude/skills/anysearch/scripts/ ~/.opencode/skills/anysearch/scripts/ ~/.openclaw/skills/anysearch/scripts/ ~/.cursor/skills/anysearch/scripts/ ~/.anysearch/scripts/ 或 which anysearch_cli.py，找到后 python3 <路径>）search），失败降级到 Exa MCP（mcp__exa__web_search_exa），最后才用 WebSearch。使用 WebSearch 时标注降级模式。"
V1 vs V2 Comparison
| Dimension | v1 (serial) | v2 (Subagent parallel) |
|---|---|---|
| Option research | Main agent searches manually | Subagent researches independently |
| Market validation | No systematic validation | Subagent validates in parallel |
| Token usage | Search results occupy the main context | Subagent processes independently |
| Basis for decision | User's subjective judgment | Objective recommendations supported by data |
Notes
- Scientific decision-making: Based on data and facts, avoid subjective assumptions
- Multi-dimensional evaluation: Comprehensively consider multiple dimensions such as finance, strategy, technology, and risk
- Dynamic adjustment: Adjust decisions promptly according to environmental changes
- Decision records: Record the decision-making process and basis for easy review
Output Quality Comparison
✅ Good example：
- 有数据引用：「根据 Q4 数据，留存率从 35% 降至 28%」
- 有验证来源：「数据来源：Google Analytics, 2025-12-01」
- 有明确建议：「建议将新手引导步骤从 5 步减少至 3 步」
❌ Bad example：
- 模糊结论：「数据表明留存率有所下降」
- 无来源：「根据经验，这个功能很重要」
- 没有行动建议：「留存是个问题」
Common Pitfalls / Red Flags — STOP
Stop immediately and backtrack if any of the following occur:
| Pitfall | Correct approach |
|---|---|
| Drawing conclusions using "should", "probably", "seems" | Must be based on actual data and validation |
| Claiming completion without running checks | Verify first, then state |
| Skipping key steps due to time pressure | No exceptions; the tighter the time, the stricter you must be |
| The thought "this time it should be fine" | Re-verify every time |
Output Quality Check / Verification Checklist
- Prerequisite dependencies are satisfied (input documents/data have been collected)
- All core steps have been executed
- Output document has been generated to the
docs/directory - Every judgment is supported by data/evidence
- 2-3 follow-up skills have been recommended
⚠️ If any item fails → complete it before marking as done.