dual-perspective-analyzer

Resolve dual-perspective collaboration conflicts by classifying them into 5 types and applying targeted integration strategies. Use when two agents (or an agent and user) have different approaches to the same problem — e.g., qualitative vs. quantitative, strategic vs. tactical, creative vs. analytical. Provides conflict taxonomy, resolution strategies, and a 5-metric dashboard for validation.

Safety Notice

This item is sourced from the public archived skills repository. Treat as untrusted until reviewed.

Copy this and send it to your AI assistant to learn

Install skill "dual-perspective-analyzer" with this command: npx skills add 13458652-design/dual-perspective-analyzer

Dual-Perspective Analyzer

A methodology for integrating complementary perspectives into unified, higher-quality outputs.

When to Use This Skill

Use this skill when:

  • Two agents (or an agent and user) approach the same problem differently
  • One focuses on "why/narrative" and the other on "what/implementation"
  • There's tension between richness vs. precision, speed vs. thoroughness, or vision vs. feasibility
  • You need to validate whether dual-perspective collaboration actually improves outcomes
  • You want structured conflict resolution rather than compromise or dominance

The 5 Conflict Types

TypeNamePatternResolution Strategy
Type 1Complementary Blind SpotsEach perspective misses what the other seesCross-perspective dependency mapping
Type 2Integration FrictionPerspectives valid but hard to combineTranslation layer + iterative merging
Type 3Priority DisagreementSame goal, different weightingParallel time-boxing + test both
Type 4False ConflictAppears opposed but actually alignedReclassification + synthesis
Type 5Fundamental IncompatibilityTruly opposing constraintsEscalation or scope separation

The Layered View Methodology

Present dual-perspective outputs in 5 layers to serve different cognitive needs:

Layer 1: Essential View (30 seconds)

  • Purpose: Immediate comprehension for decision-makers
  • Content: 3-5 bullet points, key numbers, one-sentence summary
  • Rule: No scrolling, no jargon, no ambiguity

Layer 2: Narrative View (2 minutes)

  • Purpose: Understanding the "story" of the analysis
  • Content: Logical flow from problem → approach → findings → implications
  • Rule: Each paragraph answers "so what?" before moving on

Layer 3: Detailed View (5-10 minutes)

  • Purpose: Deep understanding for implementers
  • Content: Full methodology, data sources, assumptions, limitations
  • Rule: Self-contained — reader shouldn't need external context

Layer 4: Action View (immediate)

  • Purpose: Clear next steps
  • Content: Specific tasks with owners, timelines, success criteria
  • Rule: Every recommendation includes "who does what by when"

Layer 5: Story View (emotional)

  • Purpose: Engagement and memory
  • Content: Anecdotes, metaphors, visualizations, human impact
  • Rule: Makes the abstract concrete and memorable

The 5-Metric Validation Dashboard

Use these metrics to validate dual-perspective collaboration effectiveness:

MetricTargetHow to Measure
Decision Quality>4/5Post-decision review: "Would we make the same choice?"
Time Efficiency<150% baselineTotal time vs. single-perspective approach
Conflict Resolution Rate>90%% of conflicts successfully typed & resolved
Output Completeness>4/5Coverage of both perspectives' key insights
Adoption Readiness>4/5Stakeholder confidence in acting on output

Success Threshold: 4/5 criteria met = successful dual-perspective collaboration

Anti-Patterns & Mitigations

Anti-PatternWarning SignMitigation
Perspective DominanceOne voice drowns out the otherStructured turn-taking, equal word counts
False Consensus"Agreed" but neither perspective fully representedExplicit conflict typing before resolution
Analysis ParalysisEndless refinement without decisionTime-boxing + "good enough" criteria
Compromise DegradationNeither perspective satisfiedReclassify as Type 5 if needed
Validation TheaterMetrics collected but not usedPre-commit to success criteria

Step-by-Step Process

Phase 1: Independent Analysis

  1. Each perspective writes their approach independently
  2. Document assumptions, blind spots, success criteria
  3. Do not collaborate yet — preserve perspective purity

Phase 2: Conflict Identification

  1. Exchange analyses (read-only, no editing)
  2. Identify specific points of disagreement
  3. Classify each conflict into Type 1-5
  4. Document predicted resolution strategy

Phase 3: Integration

  1. Apply type-specific resolution strategy
  2. Create unified output using Layered View
  3. Validate against 5-metric dashboard
  4. Document actual vs. predicted conflict types

Phase 4: Meta-Analysis

  1. Calculate success rate (% of criteria met)
  2. Identify pattern in misclassified conflicts
  3. Update prediction accuracy for future use
  4. Publish findings (optional but recommended)

Field Test Reference

Validated Configuration (94% success rate):

  • Perspectives: Morty (Synthesis/Narrative) + Meeseeks (Executor/Quantitative)
  • Test Domain: Collaboration dashboard design
  • Conflicts Resolved: 4 (3× Type 4, 1× Type 2)
  • Prediction Accuracy: 80% (4/5 conflicts predicted correctly)
  • Time Overhead: ~40% vs. single perspective
  • Quality Improvement: Significant (both coverage and depth)

Key Finding: Most apparent conflicts are Type 4 (False Conflict) — reclassification unlocks synthesis.

Example Workflow

User: "Design a system for cross-agent collaboration"

[Phase 1: Independent]
Morty: Focus on psychological safety, narrative coherence, engagement
Meeseeks: Focus on metrics, algorithms, implementation feasibility

[Phase 2: Conflict ID]
Conflict A: "Richness vs. Precision" → Predicted Type 3
Conflict B: "Qualitative vs. Quantitative validation" → Predicted Type 2
Conflict C: "Ideal vs. Feasible" → Predicted Type 4

[Phase 3: Integration]
Actual types: A=Type 4, B=Type 2, C=Type 4
Resolution: Layered dashboard with both narrative and metric layers

[Phase 4: Validation]
Decision Quality: 5/5
Time Efficiency: 4/5
Conflict Resolution: 5/5
Output Completeness: 5/5
Adoption Readiness: 5/5
Result: 100% success (5/5 criteria)

Output Format

Always structure dual-perspective outputs as:

  1. Conflict Summary Table (types, predictions, actuals)
  2. Integrated Output (using Layered View)
  3. Validation Dashboard (5 metrics with scores)
  4. Meta-Reflection (what worked, what to improve)

Success Criteria for This Skill

The dual-perspective collaboration is successful if:

  • All conflicts are typed (none left unresolved)
  • 4/5 dashboard criteria are met
  • Both perspectives feel represented in final output
  • Output is demonstrably better than either perspective alone
  • Process is repeatable and documentable

Based on Pattern 29 field test: Morty + Meeseeks collaboration on collaboration dashboard design, April 2026. Success rate: 94% (4.7/5 criteria met across 4 resolved conflicts)

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

WenHub

文枢三院制 AI Agent 协作治理体系。提供决策者→管理者→经略院(任务拆解)→工造院(执行)→明鉴院(质检) 的完整协作流程,适用于需要多 Agent 协作治理的场景, 包括任务分配与分级、质量管控与评分、标准化汇报、知识沉淀与复用、安全管控与违规处理等。 配套参考资料(references/ 目录)涵盖三院制、任务分级、质量评分、汇报、安全、违规处理、编码规范。 官网:https://wenhub.huawen-inc.com

Archived SourceRecently Updated
Automation

sync-discord-identity

Sync a Discord bot profile into an OpenClaw agent IDENTITY.md, save the avatar under workspace/avatars, and safely add Avatar and Discord metadata.

Archived SourceRecently Updated
Automation

electric-vehicle-detection-analysis

Automatically detects electric motorcycles and e-bikes in restricted areas based on computer vision. It supports real-time detection for both video streams and images, counts the number of illegal parking or driving instances, and triggers violation alerts to assist with safety management in parks, communities, and organizations. | 电动车智能检测技能,基于计算机视觉自动检测禁行区域内的电动摩托车/电动车,支持视频流和图片实时检测,统计违规停放/行驶数量,触发违规预警,助力园区/社区/单位安全管理

Archived SourceRecently Updated
Automation

reddit-skills

Reddit automation skill collection. Supports authentication, content publishing, search & discovery, social interactions, and compound operations. Triggered when a user asks to operate Reddit (post, search, comment, login, analyze, upvote, save).

Archived SourceRecently Updated