stakeholder-weighted-ranking
$
npx mdskill add yogsoth-ai/de-anthropocentric-research-engine/stakeholder-weighted-rankingRanks research gaps by stakeholder perspectives using weighted consensus
- Solves multi-stakeholder prioritization conflicts in research planning
- Uses context-management and subagent-spawning to model different viewpoints
- Applies AHP weighting and Borda count to merge divergent priorities
- Produces consensus rankings and stakeholder-specific priority views
SKILL.md
.github/skills/stakeholder-weighted-rankingView on GitHub ↗
---
name: stakeholder-weighted-ranking
description: "Strategy: 按利益相关者视角加权——同一 gap 在不同视角下权重不同,最终取共识排序"
version: 1.0.0
category: hypothesis-formation
type: strategy
campaign: gap-prioritization
tactics:
- scoring-matrix-construction
- priority-sensitivity-testing
sops:
- importance-scoring
- feasibility-scoring
- novelty-scoring
- impact-scoring
- ahp-weighting
- priority-synthesis
dependencies:
skills:
- context-management
- subagent-spawning
---
# Stakeholder-Weighted Ranking
按利益相关者视角加权排序:识别所有相关方(研究者、工程师、政策制定者、终端用户等),为每类相关方构建独立的权重向量,分别排序后取共识。
## 适用场景
- 研究涉及多方利益相关者(如医疗 AI:临床医生 + 患者 + 监管机构)
- 不同相关方对"重要性"的定义存在根本分歧
- 需要在多方之间建立共识或展示不同视角的排序差异
- 资助机构或合作方需要看到自己视角下的优先级
## 思维框架
**核心原则**:没有客观的"最重要 gap",只有"对谁最重要"。
流程分三层:
**第一层:相关方识别**
列出所有会受到研究结果影响的群体。每类相关方有不同的价值函数——工程师重视可行性,政策制定者重视影响力,学术研究者重视新颖性。
**第二层:视角内排序**
对每类相关方,使用与 multi-criteria-ranking 相同的四维度评分,但权重向量不同。例如:
- 学术研究者:新颖性 0.40、重要性 0.30、影响力 0.20、可行性 0.10
- 工程师:可行性 0.40、影响力 0.30、重要性 0.20、新颖性 0.10
- 政策制定者:影响力 0.45、重要性 0.35、可行性 0.15、新颖性 0.05
**第三层:共识合并**
Borda count 或加权平均各视角排序,识别"跨视角稳健的 top gap"(所有相关方都认为重要)和"视角分歧 gap"(某些相关方高度重视,其他人不在意)。
**关键洞察**:视角分歧本身是信息——分歧大的 gap 可能需要先做利益对齐,而非直接攻击。
## Budget Gate
| Tier | Gap 数量 | 相关方数量 | 共识方法 | 最终产出 |
|------|---------|-----------|---------|---------|
| S | 5–10 | 2–3 类 | 简单平均 | 各视角排序 + 共识 top-3 |
| M | 11–20 | 3–5 类 | Borda count | 各视角排序 + 共识 top-5 + 分歧分析 |
| L | 20+ | 5+ 类 | 加权 Borda + 敏感性 | 完整视角矩阵 + 共识排序 + 分歧热图 |
## 默认参考流
1. 调用 `gap-normalization` SOP:统一 gap 格式
2. 识别相关方类别(CC 自主判断或用户指定)
3. 为每类相关方调用 `ahp-weighting` SOP:生成该视角的权重向量
4. 对每类相关方并行执行四维度评分(`importance-scoring`、`feasibility-scoring`、`novelty-scoring`、`impact-scoring`)
5. 调用 `scoring-matrix-construction` tactic:构建 gap × 相关方 × 维度 三维矩阵
6. 调用 `priority-sensitivity-testing` tactic:检验相关方权重变化对共识排序的影响
7. 调用 `priority-synthesis` SOP:Borda count 合并 → 共识排序 + 分歧报告
## context-checkpoint
每轮结束后记录:
- 相关方列表及其权重向量
- 各相关方视角下的 gap 排序
- 共识排序(Borda 得分)
- 高分歧 gap 列表(标注分歧来源)
More from yogsoth-ai/de-anthropocentric-research-engine
- abductive-hypothesis-generationStrategy: 面对异常的最佳解释推理
- ablation-brainstormRemove components one by one, observe system changes to reveal hidden dependencies and generate ideas from structural gaps.
- ablation-component-mappingMap system architecture to ablatable units for ablation studies
- ablation-designDesign ablation studies to isolate component contributions in ML systems
- ablation-executionRemove components one by one from a system, record the response/impact of each removal.
- abp-vulnerability-classificationClassify assumptions on 2 axes — load-bearing (how much conclusion depends on it) × vulnerable (how likely to be false). Focuses attention on High-Load × High-Vulnerable quadrant.
- abstraction-extractionExtract abstract principles from concrete domain cases. Strips domain-specific details to reveal transferable mechanisms.
- abstraction-ladderPerform bisociation at multiple abstraction levels
- abstraction-ladderingMove between concrete and abstract framings — 3 levels up (Why?) and 3 levels down (How?) to find the most productive research level.
- abstraction-to-designAbstract biological principle to design principle. Bridge from biology to engineering.