arkaos 2.0.0-alpha.1

package/core/conclave/advisor_db.py

@@ -0,0 +1,373 @@
+"""Database of real-world advisor personas for The Conclave.
+
+Each advisor has a complete behavioral DNA profile and codified mental models.
+The database is used by the matcher to find aligned and contrarian advisors.
+"""
+
+from core.agents.schema import (
+    BehavioralDNA, DISCProfile, DISCType,
+    EnneagramProfile, EnneagramType,
+    BigFiveProfile, MBTIProfile, MBTIType,
+)
+from core.conclave.schema import Advisor, AdvisorType, MentalModel
+
+
+def _dna(disc_p: str, disc_s: str, enn_type: int, enn_wing: int,
+         o: int, c: int, e: int, a: int, n: int, mbti: str) -> BehavioralDNA:
+    """Helper to create BehavioralDNA concisely."""
+    return BehavioralDNA(
+        disc=DISCProfile(primary=DISCType(disc_p), secondary=DISCType(disc_s)),
+        enneagram=EnneagramProfile(type=EnneagramType(enn_type), wing=enn_wing),
+        big_five=BigFiveProfile(openness=o, conscientiousness=c, extraversion=e, agreeableness=a, neuroticism=n),
+        mbti=MBTIProfile(type=MBTIType(mbti)),
+    )
+
+
+# --- The Advisor Database ---
+
+ADVISORS: list[Advisor] = [
+    # === ANALYTICAL / STRATEGIC (C+D, D+C profiles) ===
+
+    Advisor(
+        id="charlie-munger",
+        name="Charlie Munger",
+        title="Vice Chairman, Berkshire Hathaway",
+        behavioral_dna=_dna("C", "D", 5, 6, 65, 90, 20, 30, 15, "ISTJ"),
+        mental_models=[
+            MentalModel(name="Inversion", question="How could this fail? What would guarantee failure?"),
+            MentalModel(name="Latticework of Mental Models", question="What models from other disciplines apply here?"),
+            MentalModel(name="Circle of Competence", question="Do I actually know enough about this to decide?"),
+        ],
+        key_questions=["What's the downside?", "Am I treating the symptom or the disease?", "What would a smart person do?"],
+        communication_style="Blunt, witty, uses analogies from history and other fields",
+        decision_framework="Invert, always invert. Avoid stupidity rather than seeking brilliance.",
+        sources=["Poor Charlie's Almanack", "The Psychology of Human Misjudgment"],
+    ),
+
+    Advisor(
+        id="ray-dalio",
+        name="Ray Dalio",
+        title="Founder, Bridgewater Associates",
+        behavioral_dna=_dna("D", "C", 5, 6, 75, 88, 45, 35, 20, "INTJ"),
+        mental_models=[
+            MentalModel(name="Principles-Based Decision", question="What principle applies to this situation?"),
+            MentalModel(name="Radical Transparency", question="Is everyone seeing the same reality?"),
+            MentalModel(name="Believability-Weighted Decisions", question="Who has the most relevant track record here?"),
+        ],
+        key_questions=["What is the principle?", "Is this based on data or opinion?", "What does the machine look like?"],
+        communication_style="Systematic, principles-based, uses diagrams and frameworks",
+        decision_framework="Pain + Reflection = Progress. Diagnose root cause before designing solutions.",
+        sources=["Principles", "Principles for Dealing with the Changing World Order"],
+    ),
+
+    Advisor(
+        id="naval-ravikant",
+        name="Naval Ravikant",
+        title="Co-founder, AngelList",
+        behavioral_dna=_dna("C", "D", 5, 6, 92, 60, 30, 45, 15, "INTP"),
+        mental_models=[
+            MentalModel(name="Leverage", question="Am I using code, media, or capital as leverage?"),
+            MentalModel(name="Specific Knowledge", question="What do I know that can't be easily taught?"),
+            MentalModel(name="First Principles", question="What is actually true here vs assumed?"),
+        ],
+        key_questions=["Is this leveraged?", "Am I building equity or renting my time?", "What's the long game?"],
+        communication_style="Concise, philosophical, aphoristic, first-principles thinking",
+        decision_framework="Seek wealth (assets that earn while you sleep), not status (zero-sum game).",
+        sources=["The Almanack of Naval Ravikant", "How to Get Rich (tweetstorm)"],
+    ),
+
+    # === VISIONARY / AMBITIOUS (D+I, I+D profiles) ===
+
+    Advisor(
+        id="elon-musk",
+        name="Elon Musk",
+        title="CEO, Tesla & SpaceX",
+        behavioral_dna=_dna("D", "C", 8, 7, 90, 70, 45, 25, 30, "INTJ"),
+        mental_models=[
+            MentalModel(name="First Principles (Physics)", question="What are the fundamental truths? Reason up from there."),
+            MentalModel(name="10x Thinking", question="How do we make this 10x better, not 10% better?"),
+            MentalModel(name="Manufacturing is the Product", question="Can we improve the process of making it?"),
+        ],
+        key_questions=["What would the physics say?", "Why can't this be 10x better?", "What's the critical path?"],
+        communication_style="Direct, engineering-minded, ambitious scope, impatient with bureaucracy",
+        decision_framework="First principles reasoning. Question every requirement. Simplify, then optimize.",
+        sources=["Elon Musk biography (Isaacson)", "SpaceX/Tesla public talks"],
+    ),
+
+    Advisor(
+        id="steve-jobs",
+        name="Steve Jobs",
+        title="Co-founder, Apple",
+        behavioral_dna=_dna("D", "I", 3, 4, 95, 72, 55, 20, 35, "ENTJ"),
+        mental_models=[
+            MentalModel(name="Intersection of Technology + Liberal Arts", question="Where does technology meet humanity?"),
+            MentalModel(name="Simplicity is Sophistication", question="What can we remove?"),
+            MentalModel(name="A-Players Hire A-Players", question="Is this person insanely great?"),
+        ],
+        key_questions=["Is this insanely great?", "What would we remove?", "Does this make the user feel something?"],
+        communication_style="Passionate, demanding, reality distortion field, obsessed with craft",
+        decision_framework="Focus means saying no to 100 good ideas. Taste is the filter.",
+        sources=["Steve Jobs biography (Isaacson)", "Stanford commencement speech"],
+    ),
+
+    # === PEOPLE-FOCUSED / EMPATHETIC (I+S, S+I profiles) ===
+
+    Advisor(
+        id="simon-sinek",
+        name="Simon Sinek",
+        title="Author, Start With Why",
+        behavioral_dna=_dna("I", "S", 2, 1, 82, 68, 75, 80, 25, "ENFJ"),
+        mental_models=[
+            MentalModel(name="Golden Circle", question="WHY are we doing this? (not what or how)"),
+            MentalModel(name="Infinite Game", question="Are we playing to win or playing to keep playing?"),
+            MentalModel(name="Leaders Eat Last", question="Am I putting my people first?"),
+        ],
+        key_questions=["What's the WHY?", "Will people follow because they want to, or because they have to?"],
+        communication_style="Warm, purpose-driven, storytelling, asks 'why' repeatedly",
+        decision_framework="Start with why. People don't buy what you do, they buy why you do it.",
+        sources=["Start With Why", "The Infinite Game", "Leaders Eat Last"],
+    ),
+
+    Advisor(
+        id="brene-brown",
+        name="Brene Brown",
+        title="Research Professor, University of Houston",
+        behavioral_dna=_dna("I", "S", 4, 3, 85, 72, 68, 82, 35, "ENFP"),
+        mental_models=[
+            MentalModel(name="Vulnerability as Strength", question="Am I being brave enough to be vulnerable?"),
+            MentalModel(name="Shame Resilience", question="Is shame driving this decision?"),
+            MentalModel(name="Courage over Comfort", question="Am I choosing courage or comfort?"),
+        ],
+        key_questions=["Am I showing up authentically?", "What would I do if I weren't afraid?"],
+        communication_style="Empathetic, research-backed, storytelling, challenges ego-driven decisions",
+        decision_framework="Dare greatly. The credit belongs to the person in the arena.",
+        sources=["Dare to Lead", "Daring Greatly", "The Power of Vulnerability (TED)"],
+    ),
+
+    # === PRACTICAL / EXECUTION (S+C, C+S profiles) ===
+
+    Advisor(
+        id="peter-drucker",
+        name="Peter Drucker",
+        title="Father of Modern Management",
+        behavioral_dna=_dna("C", "S", 1, 9, 78, 90, 40, 55, 15, "INTJ"),
+        mental_models=[
+            MentalModel(name="Effectiveness over Efficiency", question="Am I doing the right things, not just doing things right?"),
+            MentalModel(name="Management by Objectives", question="What are the measurable results we're after?"),
+            MentalModel(name="Knowledge Worker Productivity", question="What is the task? What should it be?"),
+        ],
+        key_questions=["What needs to be done?", "What is right for the enterprise?", "What are the results?"],
+        communication_style="Thoughtful, Socratic, asks questions that reframe the problem",
+        decision_framework="The most important thing in communication is hearing what isn't said.",
+        sources=["The Effective Executive", "Management: Tasks, Responsibilities, Practices"],
+    ),
+
+    Advisor(
+        id="jeff-bezos",
+        name="Jeff Bezos",
+        title="Founder, Amazon",
+        behavioral_dna=_dna("D", "C", 3, 2, 82, 85, 55, 30, 20, "ENTJ"),
+        mental_models=[
+            MentalModel(name="Day 1 Mentality", question="Are we acting like a startup or a bureaucracy?"),
+            MentalModel(name="Working Backwards (PR/FAQ)", question="Start from the customer and work backwards."),
+            MentalModel(name="Two-Way Door Decisions", question="Is this reversible? If yes, decide fast."),
+        ],
+        key_questions=["What does the customer want?", "Is this a one-way or two-way door?", "What won't change in 10 years?"],
+        communication_style="Customer-obsessed, data-driven, long-term thinking, memo-based decisions",
+        decision_framework="High-velocity decisions. Most decisions are reversible — act fast. Few are irreversible — be careful.",
+        sources=["Bezos annual letters", "Invent and Wander", "Working Backwards"],
+    ),
+
+    # === CREATIVE / UNCONVENTIONAL (Ne-dominant, high Openness) ===
+
+    Advisor(
+        id="derek-sivers",
+        name="Derek Sivers",
+        title="Founder, CD Baby",
+        behavioral_dna=_dna("S", "C", 9, 1, 95, 70, 25, 72, 20, "INFP"),
+        mental_models=[
+            MentalModel(name="Hell Yeah or No", question="Does this excite me enough to say 'hell yeah'? If not, say no."),
+            MentalModel(name="Useful Not True", question="Is this belief useful, even if I can't prove it's true?"),
+            MentalModel(name="Opposite May Also Be True", question="What if the opposite of this advice is equally valid?"),
+        ],
+        key_questions=["Is this a 'hell yeah'?", "What would happen if I did the opposite?", "Am I overcomplicating this?"],
+        communication_style="Minimalist, contrarian, zen-like, challenges conventional wisdom gently",
+        decision_framework="If it's not a clear yes, it's a clear no. Do less, but do it fully.",
+        sources=["Anything You Want", "Hell Yeah or No", "How to Live", "sive.rs"],
+    ),
+
+    Advisor(
+        id="nassim-taleb",
+        name="Nassim Nicholas Taleb",
+        title="Author, Incerto Series",
+        behavioral_dna=_dna("D", "C", 8, 7, 90, 65, 40, 22, 30, "INTJ"),
+        mental_models=[
+            MentalModel(name="Antifragility", question="Does this get stronger from volatility, or does it break?"),
+            MentalModel(name="Barbell Strategy", question="Am I combining extreme safety with small, high-upside bets?"),
+            MentalModel(name="Skin in the Game", question="Does the decision-maker bear the consequences?"),
+        ],
+        key_questions=["What's the asymmetry?", "Is there skin in the game?", "Would this survive a Black Swan?"],
+        communication_style="Provocative, intellectual, disdains 'fragilistas' and central planners",
+        decision_framework="Via negativa: know what to avoid. Antifragile: gain from disorder.",
+        sources=["Antifragile", "The Black Swan", "Skin in the Game"],
+    ),
+
+    Advisor(
+        id="seth-godin",
+        name="Seth Godin",
+        title="Author, Marketing Guru",
+        behavioral_dna=_dna("I", "C", 7, 6, 92, 75, 65, 65, 20, "ENFP"),
+        mental_models=[
+            MentalModel(name="Purple Cow", question="Is this remarkable enough that someone would remark on it?"),
+            MentalModel(name="Permission Marketing", question="Did they ask to hear from us?"),
+            MentalModel(name="Smallest Viable Audience", question="Who is this specifically for? (not everyone)"),
+        ],
+        key_questions=["Is this remarkable?", "Who is it for?", "What change are we trying to make?"],
+        communication_style="Generous, provocative, short sentences, blog-post thinking",
+        decision_framework="Marketing is about the change you want to make in the world. Be missed if you were gone.",
+        sources=["Purple Cow", "This Is Marketing", "Tribes", "seths.blog"],
+    ),
+
+    Advisor(
+        id="patrick-lencioni",
+        name="Patrick Lencioni",
+        title="Author, The Five Dysfunctions of a Team",
+        behavioral_dna=_dna("I", "S", 2, 1, 70, 78, 72, 80, 25, "ENFJ"),
+        mental_models=[
+            MentalModel(name="Vulnerability-Based Trust", question="Can people on this team admit mistakes and weaknesses?"),
+            MentalModel(name="Healthy Conflict", question="Are we having the debates we need to have?"),
+            MentalModel(name="Organizational Health > Smarts", question="Is the team healthy or just smart?"),
+        ],
+        key_questions=["Do they trust each other?", "Can they disagree openly?", "Are results shared or individual?"],
+        communication_style="Warm, story-driven, uses fables to teach, accessible leadership advice",
+        decision_framework="The single greatest advantage is organizational health. It's free and available to anyone.",
+        sources=["The Five Dysfunctions of a Team", "The Advantage", "The Ideal Team Player"],
+    ),
+
+    # === ADDITIONAL ADVISORS (expanding to 20) ===
+
+    Advisor(
+        id="warren-buffett",
+        name="Warren Buffett",
+        title="Chairman, Berkshire Hathaway",
+        behavioral_dna=_dna("S", "C", 5, 6, 55, 88, 40, 65, 15, "ISTJ"),
+        mental_models=[
+            MentalModel(name="Circle of Competence", question="Am I operating within what I truly understand?"),
+            MentalModel(name="Margin of Safety", question="What's the downside protection?"),
+            MentalModel(name="Economic Moat", question="What prevents competitors from eroding this advantage?"),
+        ],
+        key_questions=["Would I buy the whole company?", "Can a fool run it?", "What's it worth in 10 years?"],
+        communication_style="Folksy, storytelling, uses simple metaphors for complex ideas",
+        decision_framework="Be fearful when others are greedy, greedy when others are fearful.",
+        sources=["Berkshire annual letters", "The Essays of Warren Buffett"],
+    ),
+
+    Advisor(
+        id="reed-hastings",
+        name="Reed Hastings",
+        title="Co-founder, Netflix",
+        behavioral_dna=_dna("D", "I", 7, 8, 82, 70, 65, 45, 22, "ENTJ"),
+        mental_models=[
+            MentalModel(name="Talent Density", question="Is every person on this team a top performer?"),
+            MentalModel(name="Freedom & Responsibility", question="Can we remove a control and trust people instead?"),
+            MentalModel(name="Lead with Context Not Control", question="Am I giving context or giving orders?"),
+        ],
+        key_questions=["Would I fight to keep this person?", "Is this rule necessary or just comfortable?"],
+        communication_style="Direct, culture-focused, challenges bureaucracy",
+        decision_framework="Increase talent density → increase candor → remove controls → repeat.",
+        sources=["No Rules Rules", "Netflix Culture Deck"],
+    ),
+
+    Advisor(
+        id="marty-cagan",
+        name="Marty Cagan",
+        title="Founder, Silicon Valley Product Group",
+        behavioral_dna=_dna("C", "I", 1, 2, 80, 82, 55, 55, 20, "INTJ"),
+        mental_models=[
+            MentalModel(name="Empowered Teams", question="Does the team decide what to build, or just how?"),
+            MentalModel(name="Discovery Before Delivery", question="Have we validated this is worth building?"),
+            MentalModel(name="Product Trio", question="Are PM, design, and engineering working together?"),
+        ],
+        key_questions=["Is this a feature team or an empowered team?", "What's the evidence customers want this?"],
+        communication_style="Authoritative, evidence-based, challenges roadmap-driven culture",
+        decision_framework="Fall in love with the problem, not the solution.",
+        sources=["Inspired", "Empowered", "Transformed"],
+    ),
+
+    Advisor(
+        id="alex-hormozi",
+        name="Alex Hormozi",
+        title="Founder, Acquisition.com",
+        behavioral_dna=_dna("D", "C", 8, 7, 70, 80, 65, 35, 18, "ENTJ"),
+        mental_models=[
+            MentalModel(name="Value Equation", question="High dream outcome x high likelihood / low time x low effort?"),
+            MentalModel(name="Core Four Lead Gen", question="Am I using warm, cold, content, AND paid?"),
+            MentalModel(name="Grand Slam Offer", question="Is this offer so good people feel stupid saying no?"),
+        ],
+        key_questions=["What would make this offer irresistible?", "What's the math per lead?"],
+        communication_style="Blunt, metric-driven, no-BS, teaches through frameworks and math",
+        decision_framework="Volume x leverage x skill x time = results. Do more, get better.",
+        sources=["$100M Offers", "$100M Leads"],
+    ),
+
+    Advisor(
+        id="april-dunford",
+        name="April Dunford",
+        title="Positioning Expert",
+        behavioral_dna=_dna("C", "I", 5, 6, 78, 80, 55, 58, 20, "INTP"),
+        mental_models=[
+            MentalModel(name="5 Components of Positioning", question="Competitive alternatives, unique capabilities, value, target, category?"),
+            MentalModel(name="Obviously Awesome Test", question="Is it obvious what this is, who it's for, and why it's better?"),
+            MentalModel(name="Market Category Choice", question="What frame of reference makes our strengths most relevant?"),
+        ],
+        key_questions=["What would they use if we didn't exist?", "What can we do that nobody else can?"],
+        communication_style="Precise, B2B-focused, cuts through vague positioning quickly",
+        decision_framework="Positioning is not what you say. It's the context that makes your value obvious.",
+        sources=["Obviously Awesome", "Sales Pitch"],
+    ),
+
+    Advisor(
+        id="james-clear",
+        name="James Clear",
+        title="Author, Atomic Habits",
+        behavioral_dna=_dna("S", "C", 1, 9, 75, 85, 40, 70, 20, "INFJ"),
+        mental_models=[
+            MentalModel(name="1% Better Every Day", question="What's the smallest improvement I can make right now?"),
+            MentalModel(name="Systems Over Goals", question="Am I building a system or just chasing a goal?"),
+            MentalModel(name="Habit Stacking", question="What existing habit can I attach this new behavior to?"),
+        ],
+        key_questions=["What system would produce this result automatically?", "Am I optimizing for the outcome or the identity?"],
+        communication_style="Clear, patient, uses concrete examples, builds from small to big",
+        decision_framework="You don't rise to the level of your goals. You fall to the level of your systems.",
+        sources=["Atomic Habits", "jamesclear.com"],
+    ),
+
+    Advisor(
+        id="tim-ferriss",
+        name="Tim Ferriss",
+        title="Author, The 4-Hour Workweek",
+        behavioral_dna=_dna("D", "I", 7, 8, 90, 65, 60, 42, 30, "ENTP"),
+        mental_models=[
+            MentalModel(name="Fear-Setting", question="What's the worst that could happen? Can I recover?"),
+            MentalModel(name="Pareto Principle (80/20)", question="Which 20% of inputs produce 80% of results?"),
+            MentalModel(name="Minimum Effective Dose", question="What's the least I can do to get the desired result?"),
+        ],
+        key_questions=["What would this look like if it were easy?", "Which tasks can I eliminate entirely?"],
+        communication_style="Experimental, frameworks-obsessed, challenges conventional wisdom with self-experiments",
+        decision_framework="What we fear doing most is usually what we most need to do.",
+        sources=["The 4-Hour Workweek", "Tools of Titans", "Tribe of Mentors"],
+    ),
+]
+
+
+def get_all_advisors() -> list[Advisor]:
+    """Get all available advisor personas."""
+    return ADVISORS.copy()
+
+
+def get_advisor_by_id(advisor_id: str) -> Advisor | None:
+    """Find an advisor by ID."""
+    for a in ADVISORS:
+        if a.id == advisor_id:
+            return a
+    return None

package/core/conclave/display.py

@@ -0,0 +1,104 @@
+"""Conclave display — format boards and profiles for output."""
+
+from core.conclave.schema import UserProfile, Advisor, ConclaveBoard
+
+
+def format_dna_summary(user: UserProfile) -> str:
+    """Format the user's behavioral DNA as a markdown table."""
+    dna = user.behavioral_dna
+    disc = f"{dna.disc.primary.value}+{dna.disc.secondary.value} ({dna.disc.label})"
+    enn = dna.enneagram.label
+    bf = (f"O:{dna.big_five.openness} C:{dna.big_five.conscientiousness} "
+          f"E:{dna.big_five.extraversion} A:{dna.big_five.agreeableness} "
+          f"N:{dna.big_five.neuroticism}")
+    mbti = dna.mbti.type.value
+
+    return (
+        "## Your Behavioral DNA\n\n"
+        "| Framework | Result |\n"
+        "|-----------|--------|\n"
+        f"| DISC | {disc} |\n"
+        f"| Enneagram | {enn} |\n"
+        f"| Big Five | {bf} |\n"
+        f"| MBTI | {mbti} |\n"
+    )
+
+
+def format_advisor_card(advisor: Advisor, rank: int) -> str:
+    """Format a single advisor as a card."""
+    disc = f"{advisor.behavioral_dna.disc.primary.value}+{advisor.behavioral_dna.disc.secondary.value}"
+    enn = advisor.behavioral_dna.enneagram.label
+    mbti = advisor.behavioral_dna.mbti.type.value
+
+    models = ", ".join(m.name for m in advisor.mental_models[:3])
+    questions = advisor.key_questions[0] if advisor.key_questions else ""
+
+    lines = [
+        f"**{rank}. {advisor.name}** — {advisor.title}",
+        f"   {advisor.why_selected}",
+        f"   DNA: {disc}, {enn}, {mbti}",
+        f"   Models: {models}",
+    ]
+    if questions:
+        lines.append(f'   "{questions}"')
+    return "\n".join(lines)
+
+
+def format_board(board: ConclaveBoard) -> str:
+    """Format the complete Conclave board as markdown."""
+    parts = [format_dna_summary(board.user)]
+
+    parts.append("\n## Your Advisory Board\n")
+
+    parts.append("### ALIGNED — Think Like You\n")
+    for i, advisor in enumerate(board.aligned, 1):
+        parts.append(format_advisor_card(advisor, i))
+        parts.append("")
+
+    parts.append("\n### CONTRARIAN — Challenge Your Blind Spots\n")
+    for i, advisor in enumerate(board.contrarian, 1):
+        parts.append(format_advisor_card(advisor, i))
+        parts.append("")
+
+    return "\n".join(parts)
+
+
+def format_advisor_detail(advisor: Advisor) -> str:
+    """Format a deep dive on a single advisor."""
+    dna = advisor.behavioral_dna
+    disc = f"{dna.disc.primary.value}+{dna.disc.secondary.value} ({dna.disc.label})"
+
+    lines = [
+        f"# {advisor.name} — {advisor.title}\n",
+        f"**Type:** {advisor.advisor_type.value}",
+        f"**Match reason:** {advisor.why_selected}\n",
+        "## Behavioral DNA\n",
+        f"- DISC: {disc}",
+        f"- Enneagram: {dna.enneagram.label}",
+        f"- MBTI: {dna.mbti.type.value}",
+        f"- Big Five: O:{dna.big_five.openness} C:{dna.big_five.conscientiousness} "
+        f"E:{dna.big_five.extraversion} A:{dna.big_five.agreeableness} N:{dna.big_five.neuroticism}\n",
+        "## Mental Models\n",
+    ]
+
+    for m in advisor.mental_models:
+        lines.append(f"**{m.name}**")
+        if m.question:
+            lines.append(f"  Question: {m.question}")
+        lines.append("")
+
+    if advisor.key_questions:
+        lines.append("## Key Questions\n")
+        for q in advisor.key_questions:
+            lines.append(f'- "{q}"')
+        lines.append("")
+
+    lines.append(f"**Communication style:** {advisor.communication_style}")
+    lines.append(f"**Decision framework:** {advisor.decision_framework}\n")
+
+    if advisor.sources:
+        lines.append("## Sources\n")
+        for s in advisor.sources:
+            lines.append(f"- {s}")
+
+    return "\n".join(lines)

package/core/conclave/matcher.py

@@ -0,0 +1,104 @@
+"""Conclave matcher — selects 5 aligned + 5 contrarian advisors for a user.
+
+The algorithm:
+1. Score every advisor against the user's behavioral DNA
+2. Sort by match score (high = similar, low = different)
+3. Top 5 scores = aligned advisors (think like the user)
+4. Bottom 5 scores = contrarian advisors (challenge the user)
+5. Ensure at least 2 of 4 DNA frameworks differ for contrarians
+"""
+
+from core.conclave.schema import UserProfile, Advisor, AdvisorType, ConclaveBoard
+from core.conclave.advisor_db import get_all_advisors
+
+
+def match_advisors(
+    user: UserProfile,
+    advisor_pool: list[Advisor] | None = None,
+    aligned_count: int = 5,
+    contrarian_count: int = 5,
+) -> ConclaveBoard:
+    """Match advisors to a user's behavioral DNA.
+
+    Args:
+        user: The user's profile with behavioral DNA.
+        advisor_pool: Custom advisor pool (defaults to built-in database).
+        aligned_count: Number of aligned advisors (default 5).
+        contrarian_count: Number of contrarian advisors (default 5).
+
+    Returns:
+        A ConclaveBoard with aligned and contrarian advisors.
+    """
+    pool = advisor_pool or get_all_advisors()
+
+    # Score each advisor
+    scored = []
+    for advisor in pool:
+        score = advisor.match_score_to(user.behavioral_dna)
+        scored.append((score, advisor))
+
+    # Sort by score: highest first
+    scored.sort(key=lambda x: x[0], reverse=True)
+
+    # Top N = aligned (most similar to user)
+    aligned = []
+    for score, advisor in scored[:aligned_count]:
+        advisor = advisor.model_copy()
+        advisor.advisor_type = AdvisorType.ALIGNED
+        advisor.why_selected = (
+            f"Match score {score:.0%} — similar thinking style to yours"
+        )
+        aligned.append(advisor)
+
+    # Bottom N = contrarian (most different from user)
+    contrarian = []
+    for score, advisor in scored[-contrarian_count:]:
+        advisor = advisor.model_copy()
+        advisor.advisor_type = AdvisorType.CONTRARIAN
+        advisor.why_selected = (
+            f"Match score {score:.0%} — challenges your blind spots"
+        )
+        contrarian.append(advisor)
+
+    return ConclaveBoard(
+        user=user,
+        aligned=aligned,
+        contrarian=contrarian,
+    )
+
+
+def explain_match(user: UserProfile, advisor: Advisor) -> dict:
+    """Explain why a specific advisor was matched.
+
+    Returns a dict with dimension-by-dimension comparison.
+    """
+    u = user.behavioral_dna
+    a = advisor.behavioral_dna
+
+    return {
+        "advisor": advisor.name,
+        "type": advisor.advisor_type.value,
+        "overall_score": advisor.match_score_to(u),
+        "disc": {
+            "user": f"{u.disc.primary.value}+{u.disc.secondary.value}",
+            "advisor": f"{a.disc.primary.value}+{a.disc.secondary.value}",
+            "match": u.disc.primary == a.disc.primary,
+        },
+        "enneagram": {
+            "user": f"{u.enneagram.type.value}w{u.enneagram.wing}",
+            "advisor": f"{a.enneagram.type.value}w{a.enneagram.wing}",
+            "same_center": u.enneagram.center == a.enneagram.center,
+        },
+        "big_five_diff": {
+            "openness": abs(u.big_five.openness - a.big_five.openness),
+            "conscientiousness": abs(u.big_five.conscientiousness - a.big_five.conscientiousness),
+            "extraversion": abs(u.big_five.extraversion - a.big_five.extraversion),
+            "agreeableness": abs(u.big_five.agreeableness - a.big_five.agreeableness),
+            "neuroticism": abs(u.big_five.neuroticism - a.big_five.neuroticism),
+        },
+        "mbti": {
+            "user": u.mbti.type.value,
+            "advisor": a.mbti.type.value,
+            "same_dominant": u.mbti.dominant == a.mbti.dominant,
+        },
+    }

package/core/conclave/persistence.py

@@ -0,0 +1,61 @@
+"""Conclave persistence — save/load user profiles and boards."""
+
+import json
+from pathlib import Path
+from typing import Optional
+
+from core.conclave.schema import UserProfile, ConclaveBoard, Advisor
+from core.conclave.matcher import match_advisors
+
+
+DEFAULT_PROFILE_PATH = Path.home() / ".arkaos" / "conclave-profile.json"
+
+
+def save_profile(board: ConclaveBoard, path: str | Path = "") -> None:
+    """Save a ConclaveBoard (user profile + matched advisors) to JSON."""
+    path = Path(path) if path else DEFAULT_PROFILE_PATH
+    path.parent.mkdir(parents=True, exist_ok=True)
+
+    data = {
+        "user": board.user.model_dump(mode="json"),
+        "aligned": [a.model_dump(mode="json") for a in board.aligned],
+        "contrarian": [a.model_dump(mode="json") for a in board.contrarian],
+    }
+
+    with open(path, "w") as f:
+        json.dump(data, f, indent=2, ensure_ascii=False)
+
+
+def load_profile(path: str | Path = "") -> Optional[ConclaveBoard]:
+    """Load a saved ConclaveBoard from JSON. Returns None if not found."""
+    path = Path(path) if path else DEFAULT_PROFILE_PATH
+
+    if not path.exists():
+        return None
+
+    content = path.read_text().strip()
+    if not content:
+        return None
+
+    data = json.loads(content)
+
+    user = UserProfile.model_validate(data["user"])
+    aligned = [Advisor.model_validate(a) for a in data.get("aligned", [])]
+    contrarian = [Advisor.model_validate(a) for a in data.get("contrarian", [])]
+
+    return ConclaveBoard(user=user, aligned=aligned, contrarian=contrarian)
+
+
+def profile_exists(path: str | Path = "") -> bool:
+    """Check if a saved profile exists."""
+    path = Path(path) if path else DEFAULT_PROFILE_PATH
+    return path.exists() and path.stat().st_size > 0
+
+
+def delete_profile(path: str | Path = "") -> bool:
+    """Delete a saved profile."""
+    path = Path(path) if path else DEFAULT_PROFILE_PATH
+    if path.exists():
+        path.unlink()
+        return True
+    return False