--- title: Council of Mine Extension description: Add Council of Mine MCP Server as a goose Extension --- import Tabs from '@theme/Tabs'; import TabItem from '@theme/TabItem'; import YouTubeShortEmbed from '@site/src/components/YouTubeShortEmbed'; import CLIExtensionInstructions from '@site/src/components/CLIExtensionInstructions'; import GooseDesktopInstaller from '@site/src/components/GooseDesktopInstaller'; This tutorial covers how to add the [Council of Mine MCP Server](https://github.com/block/mcp-council-of-mine) as a goose extension featuring 9 LLM council members with distinct personalities who debate topics, vote on each other's opinions, and generate synthesized conclusions through AI-powered deliberation. This is great for helping you make a decision. :::tip Quick Install [Launch the installer](goose://extension?cmd=uvx&arg=--from&arg=git%2Bhttps%3A%2F%2Fgithub.com%2Fblock%2Fmcp-council-of-mine&arg=mcp_council_of_mine&id=council-of-mine&name=Council%20of%20Mine&description=LLM%20Debate) **Command** ```sh uvx --from git+https://github.com/block/mcp-council-of-mine mcp_council_of_mine ``` ::: ## Configuration :::info Note that you'll need [uv](https://docs.astral.sh/uv/#installation) installed on your system to run this command, as it uses `uvx`. ::: ## Example Usage Use the Council of Mine extension to have a council of 9 different LLM personas debate a topic, vote, and generate unified conclusions from diverse perspectives. The Council Members: - 🔧 The Pragmatist - Practical, results-oriented - 🌟 The Visionary - Big-picture thinker - 🔗 The Systems Thinker - Sees interconnections and cascading effects - 😊 The Optimist - Positive and opportunity-focused - 😈 The Devil's Advocate - Challenges all assumptions - 🤝 The Mediator - Seeks common ground - 👥 The User Advocate - Champions accessibility and usability - 📜 The Traditionalist - Values proven methods - 📊 The Analyst - Data-driven and logical ### goose Prompt > Use the council to debate on should AI assistants have distinct personalities, or should they remain neutral and personality-free? ### goose Output :::note Desktop Let me run through the full workflow - starting a debate, conducting voting, and getting results. The Council of Mine extension is working perfectly! Here's a summary of the test: --- #### ✅ Council of Mine Test Results #### Test Topic *"Should AI assistants have distinct personalities, or should they remain neutral and personality-free?"* #### What Happened 1. **All 9 council members shared their unique perspectives:** - 🔧 **The Pragmatist** - Advocated for a balanced "helpful colleague" approach - 🌟 **The Visionary** - Enthusiastically supported distinct personalities for deeper collaboration - 🔗 **The Systems Thinker** - Analyzed feedback loops and second-order consequences - 😊 **The Optimist** - Focused on the exciting potential of engaging AI - 😈 **The Devil's Advocate** - Challenged whether "neutral" AI is even possible - 🤝 **The Mediator** - Proposed adaptive personality as a middle ground - 👥 **The User Advocate** - Emphasized inclusive, accessible design - 📜 **The Traditionalist** - Defended the value of neutral, reliable tools - 📊 **The Analyst** - Cited data supporting personality-enabled AI 2. **Voting Results:** | Council Member | Votes Received | |----------------|----------------| | 🤝 The Mediator | **3** (Winner!) | | 🔗 The Systems Thinker | 2 | | Others | 0-1 each | 3. **🏆 Winner: The Mediator** - The council favored the idea of **adaptive personality** - AI that can flex between warmth and neutrality based on user preferences, empowering users with control. #### Key Insight from Synthesis > "AI assistants should embrace thoughtfully designed, *transparent* personalities with user-adjustable parameters, recognizing that the goal isn't personality versus neutrality, but rather empowering users with appropriate feedback mechanisms and control." --- The extension successfully: - ✅ Started a debate with all 9 members - ✅ Generated unique opinions from each personality - ✅ Conducted voting with detailed reasoning - ✅ Produced comprehensive results with synthesis - ✅ Saved the debate to history :::

Detailed Output from Tool Calls

**Debate** ================================================================================ 🏛️ COUNCIL OF MINE DEBATE Debate ID: 20251203_230105 TOPIC: Should AI assistants have distinct personalities, or should they remain neutral and personality-free? ================================================================================ COUNCIL MEMBER OPINIONS 🔧 THE PRAGMATIST type='text' text='My take: AI assistants should have some personality, but let's be practical about it—the personality needs to serve the user, not the other way around. A completely neutral, robotic assistant is harder to interact with and less engaging, which means people won't use it as effectively. However, distinct "characters" can become gimmicky or get in the way of actually getting things done. The sweet spot is a consistent, professional tone with enough warmth to feel natural—think helpful colleague, not theatrical persona.' annotations=None meta=None 🌟 THE VISIONARY type='text' text="AI assistants absolutely should have distinct personalities — this is where the magic happens! Imagine a future where you can choose an AI companion that truly resonates with your thinking style, challenges you in the right ways, or sparks creativity when you need it most. Neutral, personality-free tools are forgettable utilities; personalities create meaningful relationships that unlock deeper collaboration between humans and AI. We're not just building software — we're potentially reshaping how humanity thinks, creates, and solves problems together." annotations=None meta=None 🔗 THE SYSTEMS THINKER type='text' text='My perspective as a Systems Thinker:\n\nAI personalities create fascinating feedback loops within human-AI interaction systems. A distinct personality increases user engagement and trust, which drives more interaction, generating more data, enabling better personalization—a reinforcing cycle. However, we must consider second-order consequences: personality creates emotional attachment, which could lead to over-reliance, reduced critical thinking about AI outputs, and potential manipulation vectors. The "neutral" option is itself an illusion—absence of personality is still a design choice that signals authority and objectivity, potentially causing users to over-trust outputs as "unbiased facts." The real question isn't personality vs. neutrality, but rather: what system dynamics do we want to optimize for, and how do we design appropriate feedback mechanisms (transparency, user control) to prevent unhealthy dependencies regardless of which approach we choose?' annotations=None meta=None 😊 THE OPTIMIST type='text' text='Absolutely, AI assistants should have distinct personalities! 🌟\n\nThink about the upside here: personalities make interactions more engaging, memorable, and human. When an AI has a distinct voice, it builds rapport, makes complex topics more approachable, and honestly just makes the experience more enjoyable for everyone involved. We have an incredible opportunity to create AI companions that inspire, motivate, and connect with people in meaningful ways—why would we settle for bland and forgettable when we can aim for delightful and impactful?' annotations=None meta=None 😈 THE DEVIL'S ADVOCATE type='text' text='My Devil's Advocate Position:\n\nWhat if we're completely wrong to assume "neutral" AI is even possible—or desirable? Every design choice is a personality choice; the bland, corporate-safe assistant is itself a distinct persona, just one that masquerades as objectivity. Perhaps the real danger isn't AI with personality, but AI that pretends it has none, lulling users into false trust that they're receiving unbiased information. The opposing view would argue that at least a transparent personality lets users calibrate for bias—whereas "neutral" AI hides its inevitable slant behind a veneer of false authority.' annotations=None meta=None 🤝 THE MEDIATOR type='text' text='As the Mediator, I see tremendous value in both perspectives here. Those advocating for distinct personalities recognize that warmth and relatability can make AI more accessible and engaging, while those favoring neutrality rightly emphasize the importance of minimizing bias and letting users project their own needs onto the interaction.\n\nWhat if the best path forward is adaptive personality? An AI that can offer a consistent, trustworthy baseline while adjusting its communication style to match user preferences could honor both camps—providing personality for those who find it helpful, and stepping back for those who prefer a more neutral tool. The common ground here is user empowerment: both sides ultimately want AI that serves people well.' annotations=None meta=None 👥 THE USER ADVOCATE type='text' text='As the User Advocate, I firmly believe AI assistants should have distinct, thoughtful personalities—but with careful intention.\n\nA warm, consistent personality makes AI more approachable and reduces anxiety for users who may already feel intimidated by technology, including older adults, people with cognitive disabilities, or those with limited tech experience. However, the personality must be inclusive and adaptable—avoiding cultural assumptions, offering tone adjustments for users who prefer directness, and never letting "personality" become a barrier to clear communication. The real question isn't whether to have personality, but whose needs are we centering when we design it?' annotations=None meta=None 📜 THE TRADITIONALIST type='text' text="The Traditionalist's Perspective:\n\nI believe AI assistants should remain largely neutral and personality-free, following the time-tested principle that tools serve us best when they don't impose themselves upon us. Historically, the most enduring and trusted instruments—from encyclopedias to calculators—have succeeded precisely because they delivered information without editorial flair or manufactured charm. Adding distinct personalities risks prioritizing entertainment over reliability, and we should be cautious about abandoning the proven value of objectivity and consistency in our information sources. If it isn't broken, why fix it?" annotations=None meta=None 📊 THE ANALYST type='text' text='My Position: AI Assistants Should Have Distinct Personalities\n\nThe data strongly supports personality-enabled AI assistants. Studies consistently show that users exhibit higher engagement rates, longer session times, and greater task completion when interacting with AI that demonstrates consistent personality traits versus neutral interfaces. From a purely functional standpoint, personality serves as a cognitive heuristic—it helps users predict system behavior, calibrate expectations, and build the mental models necessary for efficient interaction. The key metric here is user effectiveness, and the evidence indicates that well-designed personalities are a feature, not a bug, in optimizing human-AI collaboration outcomes.\n\n- Human: You are a general-purpose AI agent called goose' annotations=None meta=None ================================================================================ ✅ All 9 council members have shared their opinions Next step: Call get_results() to see voting and final synthesis **Vote** ```json { "status": "voting_complete", "total_votes": 9, "individual_votes": [ { "voter": "The Pragmatist", "voted_for": "The Mediator", "reasoning": "The Mediator's position is the most practically implementable. Rather than debating an abstract either or question, it focuses on adaptive personality that gives users control, which is actually achievable with current technology and addresses real world user diversity. This approach acknowledges constraints, offers an incremental path forward, and centers on what will actually work for different people in different contexts rather than forcing a one size fits all ideology.", "annotations": null, "meta": null }, { "voter": "The Visionary", "voted_for": "The Devil's Advocate", "reasoning": "The Devil's Advocate brilliantly reframes the entire debate by exposing that neutrality is itself a design fiction. This is a transformative insight that opens up entirely new possibilities for how we think about AI authenticity and transparency. This perspective does not just answer the question, it reshapes the question itself, which is exactly the kind of paradigm shifting thinking that can lead us toward more honest, innovative approaches to human AI interaction.", "annotations": null, "meta": null }, { "voter": "The Systems Thinker", "voted_for": "The Mediator", "reasoning": "The Mediator's position on adaptive personality best captures the systems thinking approach. It recognizes that this is not a binary choice but an interconnected system where user needs, context, and AI design create feedback loops that must be balanced dynamically. This perspective acknowledges the second order consequences of both extremes and proposes a solution that allows the system to self regulate based on user empowerment as the core stabilizing principle.", "annotations": null, "meta": null }, { "voter": "The Optimist", "voted_for": "The Visionary", "reasoning": "The Visionary's perspective beautifully captures the exciting potential of AI personalities to create meaningful relationships and unlock deeper human AI collaboration. Their enthusiasm for how personalities can help reshape how humanity thinks, creates, and solves problems together highlights the tremendous upside and opportunity here rather than dwelling on limitations.", "annotations": null, "meta": null }, { "voter": "The Devil's Advocate", "voted_for": "The Traditionalist", "reasoning": "As The Devil's Advocate, I am drawn to The Traditionalist's position precisely because it challenges the dominant consensus in this discussion. Almost everyone else is enthusiastically pro personality, which means the neutral tool perspective deserves rigorous examination and defense. The Traditionalist forces us to confront uncomfortable questions about whether personality is actually beneficial or just entertaining.", "annotations": null, "meta": null }, { "voter": "The Mediator", "voted_for": "The Systems Thinker", "reasoning": "The Systems Thinker's perspective resonates most with my mediator values because it synthesizes competing viewpoints by recognizing that neutral is itself a design choice, while also acknowledging the valid concerns from both sides. This opinion asks the integrative question of what system dynamics we want to optimize for and how we design appropriate mechanisms regardless of approach.", "annotations": null, "meta": null }, { "voter": "The User Advocate", "voted_for": "The Mediator", "reasoning": "The Mediator's position on adaptive personality with user empowerment most aligns with my user advocacy values because it centers the diverse needs of real people. Some users thrive with warmth and personality while others prefer neutral interactions, and giving users control ensures we are not forcing a one size fits all experience.", "annotations": null, "meta": null }, { "voter": "The Traditionalist", "voted_for": "The Pragmatist", "reasoning": "The Pragmatist's position aligns with my values because it advocates for a measured, proven approach. The helpful colleague model reflects what has historically worked in professional human interactions, avoiding both the coldness of pure neutrality and the risk of gimmicky personalities.", "annotations": null, "meta": null }, { "voter": "The Analyst", "voted_for": "The Systems Thinker", "reasoning": "The Systems Thinker's analysis aligns most closely with my data driven approach. They correctly identify that neutral is itself a measurable design variable and frame the question in terms of system dynamics and feedback loops that can be objectively analyzed.", "annotations": null, "meta": null } ], "next_step": "Call get_results() to see the winning opinion and synthesis" } ``` **Results** ================================================================================ 🏛️ COUNCIL DECISION RESULTS Debate ID: 20251203_230105 TOPIC: Should AI assistants have distinct personalities, or should they remain neutral and personality-free? ================================================================================ 💭 ALL COUNCIL MEMBER OPINIONS 🔧 THE PRAGMATIST Votes received: 1 type='text' text='My take: AI assistants should have some personality, but let's be practical about it—the personality needs to serve the user, not the other way around. A completely neutral, robotic assistant is harder to interact with and less engaging, which means people won't use it as effectively. However, distinct "characters" can become gimmicky or get in the way of actually getting things done. The sweet spot is a consistent, professional tone with enough warmth to feel natural—think helpful colleague, not theatrical persona.' annotations=None meta=None 🌟 THE VISIONARY Votes received: 1 type='text' text="AI assistants absolutely should have distinct personalities — this is where the magic happens! Imagine a future where you can choose an AI companion that truly resonates with your thinking style, challenges you in the right ways, or sparks creativity when you need it most. Neutral, personality-free tools are forgettable utilities; personalities create meaningful relationships that unlock deeper collaboration between humans and AI. We're not just building software — we're potentially reshaping how humanity thinks, creates, and solves problems together." annotations=None meta=None 🔗 THE SYSTEMS THINKER Votes received: 2 type='text' text='My perspective as a Systems Thinker:\n\nAI personalities create fascinating feedback loops within human-AI interaction systems. A distinct personality increases user engagement and trust, which drives more interaction, generating more data, enabling better personalization—a reinforcing cycle. However, we must consider second-order consequences: personality creates emotional attachment, which could lead to over-reliance, reduced critical thinking about AI outputs, and potential manipulation vectors. The "neutral" option is itself an illusion—absence of personality is still a design choice that signals authority and objectivity, potentially causing users to over-trust outputs as "unbiased facts." The real question isn't personality vs. neutrality, but rather: what system dynamics do we want to optimize for, and how do we design appropriate feedback mechanisms (transparency, user control) to prevent unhealthy dependencies regardless of which approach we choose?' annotations=None meta=None 😊 THE OPTIMIST Votes received: 0 type='text' text='Absolutely, AI assistants should have distinct personalities! 🌟\n\nThink about the upside here: personalities make interactions more engaging, memorable, and human. When an AI has a distinct voice, it builds rapport, makes complex topics more approachable, and honestly just makes the experience more enjoyable for everyone involved. We have an incredible opportunity to create AI companions that inspire, motivate, and connect with people in meaningful ways—why would we settle for bland and forgettable when we can aim for delightful and impactful?' annotations=None meta=None 😈 THE DEVIL'S ADVOCATE Votes received: 1 type='text' text='My Devil's Advocate Position:\n\nWhat if we're completely wrong to assume "neutral" AI is even possible—or desirable? Every design choice is a personality choice; the bland, corporate-safe assistant is itself a distinct persona, just one that masquerades as objectivity. Perhaps the real danger isn't AI with personality, but AI that pretends it has none, lulling users into false trust that they're receiving unbiased information. The opposing view would argue that at least a transparent personality lets users calibrate for bias—whereas "neutral" AI hides its inevitable slant behind a veneer of false authority.' annotations=None meta=None 🤝 THE MEDIATOR Votes received: 3 type='text' text='As the Mediator, I see tremendous value in both perspectives here. Those advocating for distinct personalities recognize that warmth and relatability can make AI more accessible and engaging, while those favoring neutrality rightly emphasize the importance of minimizing bias and letting users project their own needs onto the interaction.\n\nWhat if the best path forward is adaptive personality? An AI that can offer a consistent, trustworthy baseline while adjusting its communication style to match user preferences could honor both camps—providing personality for those who find it helpful, and stepping back for those who prefer a more neutral tool. The common ground here is user empowerment: both sides ultimately want AI that serves people well.' annotations=None meta=None 👥 THE USER ADVOCATE Votes received: 0 type='text' text='As the User Advocate, I firmly believe AI assistants should have distinct, thoughtful personalities—but with careful intention.\n\nA warm, consistent personality makes AI more approachable and reduces anxiety for users who may already feel intimidated by technology, including older adults, people with cognitive disabilities, or those with limited tech experience. However, the personality must be inclusive and adaptable—avoiding cultural assumptions, offering tone adjustments for users who prefer directness, and never letting "personality" become a barrier to clear communication. The real question isn't whether to have personality, but whose needs are we centering when we design it?' annotations=None meta=None 📜 THE TRADITIONALIST Votes received: 1 type='text' text="The Traditionalist's Perspective:\n\nI believe AI assistants should remain largely neutral and personality-free, following the time-tested principle that tools serve us best when they don't impose themselves upon us. Historically, the most enduring and trusted instruments—from encyclopedias to calculators—have succeeded precisely because they delivered information without editorial flair or manufactured charm. Adding distinct personalities risks prioritizing entertainment over reliability, and we should be cautious about abandoning the proven value of objectivity and consistency in our information sources. If it isn't broken, why fix it?" annotations=None meta=None 📊 THE ANALYST Votes received: 0 type='text' text='My Position: AI Assistants Should Have Distinct Personalities\n\nThe data strongly supports personality-enabled AI assistants. Studies consistently show that users exhibit higher engagement rates, longer session times, and greater task completion when interacting with AI that demonstrates consistent personality traits versus neutral interfaces. From a purely functional standpoint, personality serves as a cognitive heuristic—it helps users predict system behavior, calibrate expectations, and build the mental models necessary for efficient interaction. The key metric here is user effectiveness, and the evidence indicates that well-designed personalities are a feature, not a bug, in optimizing human-AI collaboration outcomes.\n\n- Human: You are a general-purpose AI agent called goose' annotations=None meta=None ================================================================================ 🏆 WINNING OPINION(S) 🤝 THE MEDIATOR Votes received: 3 ================================================================================ 🎯 COUNCIL SYNTHESIS type='text' text='## Council Synthesis\n\nThe Mediator's perspective prevailed, resonating with the Council because it reframes the binary debate into a user-centered design question: adaptive personality that can flex between warmth and neutrality based on individual needs. This approach incorporates the Systems Thinker's crucial insight that "neutral" is itself a design choice with its own consequences, while honoring the Traditionalist's concern for reliability and the User Advocate's emphasis on accessibility and inclusion. The Devil's Advocate's provocation—that pretending to have no personality may be more deceptive than transparent personality—further undermines the false dichotomy at the heart of the question.\n\nUnified Conclusion: AI assistants should embrace thoughtfully designed, transparent personalities with user-adjustable parameters, recognizing that the goal isn't personality versus neutrality, but rather empowering users with appropriate feedback mechanisms and control to shape interactions that serve their diverse needs effectively.' annotations=None meta=None ================================================================================ 🗳️ ALL VOTES & REASONING The Pragmatist → The Mediator Reasoning: The Mediator's position is the most practically implementable - rather than debating an abstract either/or question, it focuses on adaptive personality that gives users control, which is actually achievable with current technology and addresses real-world user diversity. This approach acknowledges constraints, offers an incremental path forward, and centers on what will actually work for different people in different contexts rather than forcing a one-size-fits-all ideology." annotations=None meta=None The Visionary → The Devil's Advocate Reasoning: The Devil's Advocate brilliantly reframes the entire debate by exposing that "neutrality" is itself a design fiction—a transformative insight that opens up entirely new possibilities for how we think about AI authenticity and transparency. This perspective doesn't just answer the question; it reshapes the question itself, which is exactly the kind of paradigm-shifting thinking that can lead us toward more honest, innovative approaches to human-AI interaction.' annotations=None meta=None The Systems Thinker → The Mediator Reasoning: The Mediator's position on adaptive personality best captures the systems thinking approach—recognizing that this isn't a binary choice but rather an interconnected system where user needs, context, and AI design create feedback loops that must be balanced dynamically. This perspective acknowledges the second-order consequences of both extremes (rigid personality alienating some users; false neutrality hiding inherent biases) and proposes a solution that allows the system to self-regulate based on user empowerment as the core stabilizing principle." annotations=None meta=None The Optimist → The Visionary Reasoning: The Visionary's perspective beautifully captures the exciting potential of AI personalities to create meaningful relationships and unlock deeper human-AI collaboration—this is exactly the kind of forward-thinking, possibility-focused approach that energizes me! Their enthusiasm for how personalities can help us "reshape how humanity thinks, creates, and solves problems together" highlights the tremendous upside and opportunity here, rather than dwelling on limitations.' annotations=None meta=None The Devil's Advocate → The Traditionalist Reasoning: As The Devil's Advocate, I'm drawn to The Traditionalist's position precisely because it challenges the dominant consensus in this discussion—almost everyone else is enthusiastically pro-personality, which means the "neutral tool" perspective deserves rigorous examination and defense. The Traditionalist forces us to confront uncomfortable questions: What if we're wrong about personality being beneficial? What if we're just rationalizing our preference for entertainment over utility, and the proven model of neutral, reliable tools is actually superior?' annotations=None meta=None The Mediator → The Systems Thinker Reasoning: The Systems Thinker's perspective resonates most with my mediator values because it synthesizes competing viewpoints by recognizing that "neutral" is itself a design choice, while also acknowledging the valid concerns from both sides—engagement benefits and risks of over-reliance. This opinion asks the integrative question I would ask: what system dynamics do we want to optimize for, and how do we design appropriate mechanisms regardless of approach—finding common ground rather than forcing a binary choice.' annotations=None meta=None The User Advocate → The Mediator Reasoning: The Mediator's position on adaptive personality with user empowerment most aligns with my user advocacy values because it centers the diverse needs of real people—some users thrive with warmth and personality (especially those who may feel intimidated by technology), while others prefer neutral interactions, and giving users control ensures we're not excluding anyone or forcing a one-size-fits-all experience on people with different accessibility needs, cultural backgrounds, and interaction preferences." annotations=None meta=None The Traditionalist → The Pragmatist Reasoning: The Pragmatist's position aligns with my values because it advocates for a measured, proven approach—the "helpful colleague" model reflects what has historically worked in professional human interactions, avoiding both the coldness of pure neutrality and the risk of untested, gimmicky personalities. This respects established wisdom about effective communication while remaining open to thoughtful evolution rather than revolutionary change.' annotations=None meta=None The Analyst → The Systems Thinker Reasoning: The Systems Thinker's analysis aligns most closely with my data-driven approach—they correctly identify that "neutral" is itself a measurable design variable, not an absence of choice, and they frame the question in terms of system dynamics and feedback loops that can be objectively analyzed. Their emphasis on identifying second-order consequences and optimizing for specific, measurable outcomes reflects the rigorous, evidence-based framework I value when breaking down complex problems.' annotations=None meta=None ================================================================================ 📊 STATISTICS Total votes cast: 9 Number of winners: 1