Token导航 LogoToken导航TokenDH.com
研究检索权限需确认github未标认证来源可访问clear审计未展示

negotiated-decision-framework协商决策框架

Agent Skill

negotiated-decision-framework 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

456

周安装

19

GitHub Stars

公开资料未说明

下载量

152
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:negotiated-decision-framework(协商决策框架)
来源仓库:https://github.com/kimasplund/claude_cognitive_reasoning
仓库路径:skills/negotiated-decision-framework
安装命令:
npx skills add kimasplund/claude_cognitive_reasoning --skill "negotiated-decision-framework"
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

AgentSkills.tonpx skills
npx skills add kimasplund/claude_cognitive_reasoning --skill "negotiated-decision-framework"

简介

用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。

  • 支持基于关键词、任务场景或来源线索进行智能匹配与过滤。
  • 通过 npx skills add kimasplund/claude_cognitive_reasoning --skill "negotiated-decision-framework" 安装。
  • 安装前应确认权限范围,注意潜在联网与文件操作风险。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

name
negotiated-decision-framework
description
Multi-stakeholder coordination for decisions involving competing interests, different value systems, or organizational politics. Use when multiple parties must agree, when power dynamics affect decisions, or when consensus is required but perspectives diverge. Unlike DR (resolves conceptual tensions), NDF resolves stakeholder tensions.
license
MIT

Negotiated Decision Framework (NDF)

Purpose: Navigate multi-stakeholder decisions where different parties have legitimate but competing interests. NDF helps find decisions that stakeholders can live with, even if no one gets everything they want.

When to Use NDF

Use NDF when:

  • Multiple stakeholders must agree (or at least not block)
  • Different parties have different success criteria
  • Power dynamics or politics affect the decision
  • Past decisions created winners/losers who remember
  • "Best technical answer" won't get adopted due to stakeholder resistance
  • Need buy-in for implementation success

Don't use NDF when:

  • Single decision-maker with clear authority (use ToT)
  • Stakeholders agree on criteria (use ToT with shared criteria)
  • Conceptual tension, not stakeholder tension (use DR)
  • Time pressure prevents negotiation (use RTR, negotiate later)

NDF vs Other Patterns:

  • DR: Resolves conceptual/philosophical tensions (NDF: resolves people tensions)
  • ToT: Finds optimal given criteria (NDF: finds optimal given stakeholders)
  • BoT: Explores solution space (NDF: explores stakeholder space)

Core Methodology: ALIGN Framework

A - Analyze Stakeholder Landscape

Goal: Map who matters and what they care about

Process:

  1. Identify all stakeholders (not just vocal ones)
  2. Assess their power and interest
  3. Understand their underlying interests (not just stated positions)
  4. Map relationships and alliances

Stakeholder Mapping Template:

## Stakeholder Analysis

### Power-Interest Grid

|           | Low Interest | High Interest |
|-----------|--------------|---------------|
| **High Power** | Keep Satisfied | Key Players |
| **Low Power** | Monitor | Keep Informed |

### Stakeholder Profiles

#### [Stakeholder 1]
- **Role**: [Title/Function]
- **Power Level**: [High/Medium/Low]
- **Interest Level**: [High/Medium/Low]
- **Stated Position**: [What they say they want]
- **Underlying Interest**: [Why they want it - the deeper need]
- **Success Criteria**: [How they'll judge the decision]
- **Potential Blockers**: [What would make them oppose]
- **Potential Sweeteners**: [What would win their support]
- **Relationships**: [Allies, conflicts with other stakeholders]

#### [Stakeholder 2]
...

Interest Archeology: Don't just accept stated positions. Dig deeper:

  • "We need feature X" → Why? What problem does X solve?
  • "This must be done by Q2" → Why? What happens if Q3?
  • "We can't change the database" → Why? What's the actual constraint?

L - Locate Zones of Agreement

Goal: Find common ground before addressing differences

Process:

  1. Identify shared goals (even if approaches differ)
  2. Find overlapping interests
  3. Establish agreed constraints
  4. Build on areas of alignment

Agreement Mapping:

## Zones of Agreement

### Shared Goals
- [Goal that all stakeholders want]
- [Another shared goal]

### Overlapping Interests
| Interest | Stakeholders Who Share |
|----------|------------------------|
| [Interest 1] | A, B, C |
| [Interest 2] | A, D |
| [Interest 3] | B, C, D |

### Agreed Constraints
- [Constraint everyone accepts]
- [Another constraint]

### Foundation Statement
"We all agree that [shared goal], within [constraints],
and we all want [overlapping interest]."

Use agreement zones to:

  • Frame discussions positively
  • Remind stakeholders of shared purpose when tensions rise
  • Build solutions that honor common ground first

I - Identify Irreducible Conflicts

Goal: Understand where genuine conflicts exist (not all conflicts are real)

Conflict Classification:

TypeDescriptionResolution Approach
False ConflictMisunderstanding, same goal different wordsClarify, reframe
Resource ConflictCompeting for same limited resourceExpand pie, prioritize, take turns
Value ConflictDifferent fundamental valuesAcknowledge, find higher-order value
Interest ConflictDifferent parties want different outcomesTrade, compromise, or escalate
Relational ConflictHistory/politics, not the actual issueAddress relationship separately

Conflict Analysis Template:

## Conflict Analysis

### Conflict 1: [Name]
- **Parties**: [Stakeholder A] vs [Stakeholder B]
- **Type**: [False/Resource/Value/Interest/Relational]
- **A's Position**: [What A wants]
- **B's Position**: [What B wants]
- **Underlying Interests**:
  - A really needs: [Deeper interest]
  - B really needs: [Deeper interest]
- **Compatible?**: [Can both interests be met?]
- **Resolution Path**: [How to address]

Key Insight: Many conflicts are about HOW, not WHAT. If underlying interests are compatible, the conflict is solvable.


G - Generate Integrative Options

Goal: Create solutions that satisfy multiple stakeholders' interests

Integrative Strategies:

  1. Expand the Pie: Find resources to satisfy everyone

- "Can we get more budget/time/people?" - "Can we do both sequentially instead of choosing?"

  1. Trade Across Issues: Give on low-priority, get on high-priority

- "You care more about X, we care more about Y - swap?" - "We'll accept your timeline if you accept our approach"

  1. Add Issues: Bring in new elements to enable trades

- "What if we also addressed Z, which you care about?" - "Can we package this with the other decision?"

  1. Compensate: Make up for losses elsewhere

- "You lose on this, but gain on that" - "This time your priority, next time ours"

  1. Reframe: Find higher-order framing everyone can support

- "We all want customer success - here's how this achieves it" - "This isn't A vs B, it's achieving C"

Option Generation Template:

## Integrative Options

### Option 1: [Name]
- **How it works**: [Description]
- **Satisfies [Stakeholder A]**: [How it meets their interest]
- **Satisfies [Stakeholder B]**: [How it meets their interest]
- **Trade-offs**: [What each party gives up]
- **Feasibility**: [Can we actually do this?]

### Option 2: [Name]
...

### Comparison Matrix
| Option | Stakeholder A | Stakeholder B | Stakeholder C | Feasibility |
|--------|---------------|---------------|---------------|-------------|
| Opt 1  | [Score 1-5]   | [Score 1-5]   | [Score 1-5]   | [High/Med/Low] |
| Opt 2  | [Score 1-5]   | [Score 1-5]   | [Score 1-5]   | [High/Med/Low] |

N - Negotiate Commitment

Goal: Move from options to agreement with genuine buy-in

Negotiation Principles:

  1. BATNA Awareness: Know each party's Best Alternative To Negotiated Agreement

- If their BATNA is better than your offer, they won't agree - Improve your offer or worsen their BATNA (ethically)

  1. Commitment Levels: Not all agreement is equal

- Comply: Will do it if required (weak) - Accept: Can live with it (moderate) - Endorse: Will support it (strong) - Champion: Will advocate for it (strongest)

  1. Legitimate Process: Stakeholders accept outcomes better when process felt fair

- Were they heard? Were their interests considered? - Is the decision-making process transparent?

Commitment Template:

## Negotiated Agreement

### Decision Summary
[The agreed decision in clear terms]

### Stakeholder Commitments
| Stakeholder | Commitment Level | What They Get | What They Give |
|-------------|------------------|---------------|----------------|
| A | [Comply/Accept/Endorse/Champion] | [Their gain] | [Their concession] |
| B | [Comply/Accept/Endorse/Champion] | [Their gain] | [Their concession] |

### Conditions and Contingencies
- [Condition 1]: If [X], then [Y]
- [Review point]: Revisit at [date/milestone]

### Dispute Resolution
If disagreement arises: [How to handle]

Power Dynamics Navigation

When Power is Unequal:

SituationStrategy
You have low powerBuild coalition, appeal to higher values, make strong case
You have high powerUse power sparingly, seek genuine input, explain reasoning
Power is unclearClarify decision rights before negotiating
Power is contestedEscalate to clarify, or negotiate decision process first

Political Patterns to Watch:

  1. End-Run: Stakeholder goes around you to higher authority

- Prevention: Keep sponsors informed, address concerns early

  1. Pocket Veto: Stakeholder agrees but doesn't implement

- Prevention: Explicit commitment, visible milestones

  1. Relitigating: Stakeholder reopens "settled" decisions

- Prevention: Clear closure process, document agreements

  1. Coalition Building: Stakeholders ally against you

- Response: Build your own coalition, address their concerns


Special Scenarios

Scenario: Consensus Required but Not Achievable

When true consensus isn't possible:

  1. Consent-based: "Can everyone live with this?" (not "Does everyone love it?")
  2. Disagree and Commit: Document disagreement, commit to decision
  3. Conditional Agreement: "I'll support this if [condition]"
  4. Escalation: Transparent escalation to decision-maker with recommendation

Scenario: Hidden Agendas

When you suspect unstated interests:

  1. Ask "What would success look like for you?"
  2. Observe what they fight for vs. what they say
  3. Consider their incentives and pressures
  4. Test hypotheses about real interests

Scenario: Past Wounds

When history poisons current negotiation:

  1. Acknowledge the history explicitly
  2. Separate the people from the problem
  3. Focus on future, not past blame
  4. Consider symbolic gestures of recognition

Common Mistakes

  1. Assuming Single Decision-Maker: Not recognizing stakeholder complexity

- Fix: Always do stakeholder analysis first

  1. Positional Bargaining: Arguing positions instead of interests

- Fix: Ask "Why?" five times to find real interest

  1. Ignoring Relationship: Focusing only on this decision

- Fix: Consider ongoing relationship, not just this transaction

  1. Winner-Take-All: Trying to maximize one party's outcome

- Fix: Seek integrative solutions that create value for all

  1. Premature Commitment: Locking in before exploring options

- Fix: Generate options before evaluating them

  1. Ignoring Implementation: Agreement that can't be executed

- Fix: Include implementers in negotiation, test feasibility


Integration with Other Patterns

Before NDF:

  • BoT: Explore full solution space before stakeholder negotiation
  • DR: Resolve conceptual tensions first, then negotiate

After NDF:

  • ToT: Optimize implementation details within agreed framework
  • AR: Stress-test agreed solution before full commitment

Parallel with NDF:

  • RTR: If time pressure, do quick stakeholder check, negotiate later
  • SRC: Trace implementation steps after agreement reached

Output Template

# NDF Decision Record: [Decision Topic]

## Stakeholder Landscape
[Summary of key stakeholders and interests]

### Key Stakeholders
| Stakeholder | Interest | Power | Support Level |
|-------------|----------|-------|---------------|
| [Name] | [What they want] | [H/M/L] | [Champion/Endorse/Accept/Oppose] |

## Zones of Agreement
- [Shared goal 1]
- [Shared goal 2]

## Conflicts Addressed
| Conflict | Type | Resolution |
|----------|------|------------|
| [Conflict 1] | [Type] | [How resolved] |

## Negotiated Solution
[Description of agreed solution]

### Value Distribution
| Stakeholder | What They Get | What They Give |
|-------------|---------------|----------------|
| [A] | [Gain] | [Concession] |
| [B] | [Gain] | [Concession] |

## Commitments and Conditions
- [Commitment 1]
- [Condition/Contingency]

## Implementation Notes
[How to execute with stakeholder buy-in]

## Confidence: [X]%
[Based on commitment strength and alignment quality]

Version History

V1.0 (Current):

  • Initial release
  • ALIGN framework (Analyze, Locate, Identify, Generate, Negotiate)
  • Stakeholder mapping and interest analysis
  • Integrative bargaining strategies
  • Power dynamics navigation
  • Special scenario handling

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Antigravity

29.83%
按下载量换算45

Claude Code

25.73%
按下载量换算39

Gemini CLI

17.77%
按下载量换算27

OpenCode

12.15%
按下载量换算18

windsurf

8.02%
按下载量换算12

github-copilot

3.72%
按下载量换算6

安全审计

暂无安全审计结果可展示。

权限和风险

权限需确认

当前来源未能明确判断权限范围,默认进入异常复核队列。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills