Skip to content
Back to skills

Multi Criteria Ranking

ASecurity

Strategy: 多维度加权评分排序——将 gap 分解为独立子问题后重组为优先级列表

  • 417 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added May 27, 2026
researchgotesting

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned May 27, 2026

npx -y skills add yogsoth-ai/de-anthropocentric-research-engine --skill multi-criteria-ranking --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Multi Criteria Ranking?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Multi Criteria Ranking
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/yogsoth-ai-multi-criteria-ranking/badge)](https://www.skillsdirectory.com/skills/yogsoth-ai-multi-criteria-ranking)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: multi-criteria-ranking
description: "Strategy: 多维度加权评分排序——将 gap 分解为独立子问题后重组为优先级列表"
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
---

# Multi-Criteria Ranking

多维度加权评分排序:将"哪个 gap 更好"这一复合问题分解为若干独立维度,分别评分后通过加权求和重组为最终排序。

## 适用场景

- gap 数量在 5–20 个之间
- 需要系统化、可解释的排序依据
- 决策者需要看到每个维度的得分(而非黑箱排序)
- 后续需要向他人解释为何选择某个 gap

## 思维框架

**核心原则**:复杂判断的可靠性来自分解,而非整体直觉。

将"哪个 gap 更值得攻击"拆解为四个独立子问题:

1. **重要性**(Importance):这个 gap 填补后,领域会前进多少?
2. **可行性**(Feasibility):以现有资源和方法,能在合理时间内解决吗?
3. **新颖性**(Novelty):这个 gap 是否真正未被充分探索?
4. **影响力**(Impact):解决后的下游效应有多广?

每个维度独立评分(1–5),避免维度间的相互污染。权重由 AHP(层次分析法)或用户指定。最终得分 = Σ(维度得分 × 维度权重)。

**敏感性检验**:权重扰动 ±20%,若排序不变则结论稳健;若排序翻转则需标记为"权重敏感"。

## Budget Gate

| Tier | Gap 数量 | 评分维度 | 敏感性检验 | 最终产出 |
|------|---------|---------|-----------|---------|
| S | 5–8 | ≥3 维度 | 可选 | 排序表 + 前 2 gap 攻击建议 |
| M | 9–15 | ≥4 维度 | 必须 | 排序表 + 前 3 gap 攻击建议 |
| L | 16–20 | ≥5 维度 | 必须(多权重场景) | 排序表 + 前 5 gap 攻击建议 + 权重敏感性报告 |

## 默认参考流

1. 调用 `gap-normalization` SOP:将输入 gaps 统一为标准格式(ID、标题、一句话描述)
2. 调用 `ahp-weighting` SOP:确定各维度权重(默认:重要性 0.35、可行性 0.25、新颖性 0.20、影响力 0.20)
3. 并行调用四个评分 SOP(`importance-scoring`、`feasibility-scoring`、`novelty-scoring`、`impact-scoring`)
4. 调用 `scoring-matrix-construction` tactic:汇总为评分矩阵
5. 调用 `priority-sensitivity-testing` tactic:扰动权重,检验排序稳健性
6. 调用 `priority-synthesis` SOP:生成最终排序 + 攻击建议

## context-checkpoint

每轮结束后记录:
- 当前评分矩阵(所有 gap × 所有维度)
- 当前权重向量
- 敏感性检验结果(稳健 / 权重敏感,标注翻转的 gap 对)
- 当前 top-N 排序

Files in this skill

  • SKILL.md2.9 KB
  • prompt.md3 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.