collab-collaborative-thinking

name	collab - Collaborative Thinking
description	Multi-stakeholder reasoning, perspective integration, finding solutions that work for multiple parties, identifying hidden agreements, and surfacing conflicting assumptions between people.
output	{"format":"prose"}

Collaborative Thinking

Input: $ARGUMENTS

---

Interpretations

Before executing, identify which interpretation matches the user's input:

Interpretation 1 — Resolve a multi-stakeholder conflict: Multiple parties disagree on a decision or direction, and the user needs to find a path that addresses the core concerns of each party without false compromise. Interpretation 2 — Surface hidden agreements and disagreements: The user senses that people are talking past each other and wants to map where parties actually agree (without knowing it) and where they genuinely conflict (instead of where they think they conflict). Interpretation 3 — Integrate perspectives into a stronger solution: The user has input from multiple people or viewpoints and wants to synthesize them into something better than any individual perspective — not averaging, but integrating.

If ambiguous, ask: "I can help with resolving a multi-party conflict, surfacing hidden agreements and real disagreements, or integrating multiple perspectives into a stronger solution — which fits?" If clear from context, proceed with the matching interpretation.

---

Core Principles

1. Positions hide interests. What people SAY they want (position) is usually not what they actually NEED (interest). Two people who hold opposing positions often share underlying interests. Finding the interest behind the position is where solutions live.

2. Disagreement is data. When smart people disagree, the disagreement itself is informative. It means they are weighting evidence differently, operating from different assumptions, or optimizing for different objectives. Do not resolve disagreement by compromise — resolve it by understanding what drives it.

3. Agreement has layers. People can agree on goals but disagree on methods. They can agree on methods but disagree on timing. They can agree on everything stated but hold conflicting unstated assumptions. Map the layer where the real agreement or disagreement lives.

4. False compromise destroys value. Splitting the difference between two positions usually gives each party half of what they want. Integrative solutions find options that give each party most of what they need by addressing different interests. Never average when you can integrate.

5. Assumptions diverge silently. Two people can use the same words, agree on the same plan, and hold completely different models of why it will work. These hidden assumption gaps are the most common source of collaboration failure — and the hardest to detect because everyone thinks they agree.

6. Power shapes what gets said. In any group, some perspectives are easier to voice than others. The most useful information often comes from the person with the least power to share it. Actively seek the perspective that is hardest to express in this context.

---

Phase 1: MAP — Identify Stakeholders and Their Views

Step 1: Stakeholder Inventory

STAKEHOLDERS:

[C1] Stakeholder: [name or role]
  Stated position: [what they say they want]
  Decision authority: [can they block, approve, or only influence?]
  Information advantage: [what do they know that others do not?]

[C2] Stakeholder: [name or role]
  Stated position: [what they say they want]
  Decision authority: [block / approve / influence]
  Information advantage: [what do they know that others do not?]

[Continue for all relevant parties]

MISSING VOICES: [Who is affected but not at the table? Whose perspective is absent?]

Step 2: Separate Positions from Interests

For each stakeholder, dig beneath the stated position:

POSITION-INTEREST MAP:

[C3] Stakeholder: [name]
  Position (what they say): [stated demand or preference]
  Interest (what they need): [underlying need the position serves]
  Fear (what they are protecting against): [what outcome they are trying to avoid]
  Constraint (what they cannot change): [non-negotiable or external limitation]

  Question to test: "If you got [position], but it did not achieve [interest], would that still be acceptable?"

[C4] Stakeholder: [name]
  Position: [stated demand]
  Interest: [underlying need]
  Fear: [what they are avoiding]
  Constraint: [non-negotiable]

[Continue for all stakeholders]

Step 3: Map Assumptions

Each stakeholder operates from assumptions about how the world works. These are often invisible and divergent.

ASSUMPTION MAP:

[C5] Topic: [specific point of apparent agreement or disagreement]

  [Stakeholder A] assumes: [their mental model]
  [Stakeholder B] assumes: [their mental model]
  [Stakeholder C] assumes: [their mental model]

  Alignment: [ALIGNED / PARTIALLY ALIGNED / CONFLICTING / UNKNOWN]
  Risk if unaddressed: [what goes wrong if this divergence is not surfaced]

[C6] Topic: [next point]
  [Continue same structure]

---

Phase 2: ANALYZE — Find the Real Structure of Agreement and Disagreement

Step 4: Agreement-Disagreement Matrix

Map where parties actually agree and disagree across multiple layers:

AGREEMENT-DISAGREEMENT MATRIX:

| Layer | [Stakeholder A] | [Stakeholder B] | [Stakeholder C] | Alignment |
|-------|-----------------|-----------------|-----------------|-----------|
| Goal (what success looks like) | [view] | [view] | [view] | [Agree/Disagree] |
| Diagnosis (what the problem is) | [view] | [view] | [view] | [Agree/Disagree] |
| Method (how to achieve it) | [view] | [view] | [view] | [Agree/Disagree] |
| Timing (when to act) | [view] | [view] | [view] | [Agree/Disagree] |
| Priority (what matters most) | [view] | [view] | [view] | [Agree/Disagree] |
| Risk tolerance (how much uncertainty is OK) | [view] | [view] | [view] | [Agree/Disagree] |
| Assumptions (what must be true) | [view] | [view] | [view] | [Agree/Disagree] |

[C7] Hidden agreements: Where do parties agree without realizing it? What goals, diagnoses, or values do they share that have been obscured by positional disagreement?

[C8] Real disagreements: Where do parties genuinely conflict at the level of interests (not just positions)? What is the actual fork — what cannot be reconciled by creative solution design?

[C9] Pseudo-disagreements: Where are parties using different words for the same thing, or arguing about method when they agree on goal?

---

Phase 3: INTEGRATE — Build Multi-Party Solutions

Step 5: Generate Integrative Options

For each real disagreement, search for options that satisfy the interests (not positions) of multiple parties:

INTEGRATIVE OPTION GENERATION:

Real disagreement: [C8 reference — the genuine conflict]

[C10] Option: [description]
  Addresses [Stakeholder A]'s interest of [interest] by: [how]
  Addresses [Stakeholder B]'s interest of [interest] by: [how]
  Trade-off: [what each party gives up — be explicit]
  Feasibility: [can this actually be implemented?]

[C11] Option: [description]
  Addresses interests by: [how]
  Trade-off: [what]
  Feasibility: [assessment]

[Continue for each option]

Rules for generating integrative options:

• Do NOT split the difference — that is compromise, not integration
• Look for differences in priorities: A cares about X, B cares about Y — give each what they care most about
• Look for differences in time horizon: A needs it now, B needs it eventually — sequence can work
• Look for differences in risk tolerance: the risk-tolerant party absorbs risk, the risk-averse party gets certainty
• Expand the pie before dividing it: are there resources, options, or framings not yet considered?

Step 6: Test Solutions Against Each Perspective

SOLUTION STRESS TEST:

Proposed solution: [description]

| Stakeholder | Interest Met? | Fear Addressed? | Constraint Respected? | Likely Response |
|-------------|---------------|-----------------|----------------------|-----------------|
| [A] | [Yes/Partial/No] | [Yes/No] | [Yes/No] | [Accept/Resist/Conditional] |
| [B] | [Yes/Partial/No] | [Yes/No] | [Yes/No] | [Accept/Resist/Conditional] |
| [C] | [Yes/Partial/No] | [Yes/No] | [Yes/No] | [Accept/Resist/Conditional] |

[C12] BLOCKING ISSUES: [What would cause any party to reject this? Can it be addressed?]
[C13] IMPLEMENTATION RISK: [Even if all agree, what could go wrong in execution due to lingering assumption gaps?]

Step 7: Surface Remaining Risks

COLLABORATION RISKS:

[C14] Assumption gap risk: [Where do parties still hold different assumptions about WHY this solution will work?]
  Mitigation: [How to surface and resolve before implementation]

[C15] Commitment asymmetry: [Who has less to lose if this fails? Are incentives aligned?]
  Mitigation: [How to equalize stakes]

[C16] Communication debt: [What has been left unsaid that will create problems later?]
  Mitigation: [What needs to be made explicit]

---

Failure Modes

Failure	Signal	Fix
Position-level negotiation	Discussing what people want without asking why they want it	Ask "what would that give you?" or "what are you trying to avoid?" for every position
Premature compromise	"Let's meet in the middle" before understanding interests	Map interests first. Compromise is a last resort, not a starting point
Loudest voice wins	Solution tracks the most powerful or articulate stakeholder	Explicitly seek the perspective that is hardest to voice in this room
False consensus	Everyone agrees verbally but holds different mental models	Test: "Tell me what you think will happen in month 3." Divergent answers reveal hidden disagreement
Interest laundering	Presenting personal preferences as objective requirements	For each "requirement," ask: who needs this and what happens if they do not get it?
Missing stakeholder	Solution works for everyone at the table but fails someone absent	Ask: who is affected but not represented? What would they say?

---

Depth Scaling

Default: 2x. Parse depth from $ARGUMENTS if specified (e.g., "/collab 4x [input]").

Depth	Min Stakeholders	Min Assumption Gaps	Min Integrative Options	Min Layers in Agreement Matrix
1x	2	2	2	3
2x	3	3	3	5
4x	5	5	5	7
8x	7	8	7	7

These are floors. Go deeper where insight is dense. Compress where it is not.

---

Pre-Completion Checklist

• All relevant stakeholders identified, including missing voices
• Positions separated from interests for every party
• Assumption gaps mapped — where do people think they agree but do not?
• Agreement-disagreement matrix completed across multiple layers
• Hidden agreements explicitly surfaced
• Real disagreements distinguished from pseudo-disagreements
• Integrative options generated (not compromise — integration)
• Solution stress-tested against each stakeholder's interests, fears, and constraints
• Remaining collaboration risks identified with mitigations

---

Integration

• Use from: /dcp (when a decision involves multiple parties), /rca (when root cause is a collaboration failure), /aex (when conflicting assumptions need extraction)
• Routes to: /reframe (when the problem framing itself is the disagreement), /prob (when parties disagree about likelihoods), /systhink (when the conflict is structural, not interpersonal)
• Differs from: /cmp compares options objectively; /collab integrates subjective perspectives from multiple stakeholders
• Complementary: /aex (extract assumptions driving each party), /reframe (find a framing that resolves apparent conflicts), /ecoc (codify shared agreements into principles)