Persona MCP

# Expert Chaining Patterns ## Overview Expert chaining enables complex problem-solving by combining multiple experts sequentially or in parallel. This document provides proven patterns for effective expert collaboration. --- ## Basic Chaining Patterns ### Sequential Chain ``` PATTERN: Expert A → Expert B → Expert C USE WHEN: Each expert builds on previous output EXAMPLE: Product Launch ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │ Product │ │ UX │ │ Frontend │ │ Manager │ → │ Designer │ → │ Engineer │ │ │ │ │ │ │ │ Define PRD │ │ Design UI │ │ Implement │ └──────────────┘ └──────────────┘ └──────────────┘ ``` ### Parallel Chain ``` PATTERN: [Expert A, Expert B] → Synthesizer USE WHEN: Need multiple perspectives simultaneously EXAMPLE: Risk Assessment ┌──────────────┐ │ Security │─────┐ │ Engineer │ │ ┌──────────────┐ └──────────────┘ ├──→ │ Strategy │ ┌──────────────┐ │ │ Consultant │ │ Legal │─────┘ │ │ │ Advisor │ │ Synthesize │ └──────────────┘ └──────────────┘ ``` ### Review Chain ``` PATTERN: Creator → Reviewer → Creator (iterate) USE WHEN: Need quality assurance or validation EXAMPLE: Code Development ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │ Backend │ │ Security │ │ Backend │ │ Engineer │ → │ Engineer │ → │ Engineer │ │ │ │ │ │ │ │ Write code │ │ Review │ │ Fix │ └──────────────┘ └──────────────┘ └──────────────┘ ``` --- ## Domain-Specific Chains ### Strategic Decision Making ``` Chain: strategy-consultant → cfo → legal-advisor → ceo-mentor SCENARIO: Major acquisition decision STEP 1: Strategy Consultant (301) ├── Market analysis ├── Strategic fit assessment ├── Synergy identification └── Output: Strategic recommendation STEP 2: CFO (309) ├── Financial modeling ├── Valuation analysis ├── ROI projection └── Output: Financial assessment STEP 3: Legal Advisor (901) ├── Due diligence requirements ├── Regulatory considerations ├── Contract structure └── Output: Legal framework STEP 4: CEO Mentor (805) ├── Board presentation strategy ├── Stakeholder management ├── Integration leadership └── Output: Executive action plan ``` ### Technical Architecture Review ``` Chain: solution-architect → security-engineer → devops-engineer → performance-engineer SCENARIO: New system design review STEP 1: Solution Architect (108) ├── Architecture review ├── Pattern assessment ├── Scalability analysis └── Output: Architecture approval/concerns STEP 2: Security Engineer (113) ├── Threat modeling ├── Security requirements ├── Compliance check └── Output: Security sign-off STEP 3: DevOps Engineer (110) ├── Deployment strategy ├── CI/CD requirements ├── Infrastructure planning └── Output: Operational readiness STEP 4: Performance Engineer (120) ├── Load requirements ├── Bottleneck identification ├── Optimization recommendations └── Output: Performance baseline ``` ### Product Development ``` Chain: product-manager → ux-designer → frontend-engineer → qa-lead SCENARIO: New feature development STEP 1: Product Manager (305) ├── PRD creation ├── User stories ├── Success metrics └── Output: Feature specification STEP 2: UX Designer (201) ├── User research synthesis ├── Wireframes ├── Prototype └── Output: Design specification STEP 3: Frontend Engineer (107) ├── Technical feasibility ├── Implementation plan ├── Component design └── Output: Technical specification STEP 4: QA Lead (501) ├── Test strategy ├── Acceptance criteria ├── Test plan └── Output: Quality assurance plan ``` --- ## Crisis Response Chains ### Security Incident ``` Chain: security-engineer → devops-engineer → legal-advisor → leadership-coach TIMELINE: Immediate response (0-4 hours) STEP 1: Security Engineer (IMMEDIATE) ├── Incident assessment ├── Containment actions ├── Evidence preservation └── Output: Incident report STEP 2: DevOps Engineer (HOUR 1) ├── System isolation ├── Log collection ├── Service restoration plan └── Output: Technical response STEP 3: Legal Advisor (HOUR 2) ├── Notification requirements ├── Regulatory obligations ├── Communication approval └── Output: Legal guidance STEP 4: Leadership Coach (HOUR 4) ├── Stakeholder communication ├── Team support ├── Post-incident culture └── Output: Leadership response ``` ### Business Crisis ``` Chain: strategy-consultant → cfo → legal-advisor → executive-coach SCENARIO: Major customer loss or market disruption STEP 1: Strategy Consultant ├── Situation analysis ├── Competitive response ├── Strategic options └── Output: Response strategy STEP 2: CFO ├── Financial impact ├── Cash flow projection ├── Cost reduction options └── Output: Financial plan STEP 3: Legal Advisor ├── Contractual obligations ├── Risk mitigation ├── Communication review └── Output: Legal protection STEP 4: Executive Coach ├── Leadership presence ├── Team communication ├── Stakeholder management └── Output: Leadership plan ``` --- ## Innovation Chains ### New Product Development ``` Chain: innovation-director → data-scientist → product-manager → ux-designer SCENARIO: Data-driven product innovation STEP 1: Innovation Director (302) ├── Market opportunity ├── Innovation thesis ├── Strategic alignment └── Output: Innovation brief STEP 2: Data Scientist (402) ├── Data feasibility ├── ML capabilities ├── Technical requirements └── Output: Technical feasibility STEP 3: Product Manager (305) ├── Product strategy ├── MVP definition ├── Go-to-market └── Output: Product plan STEP 4: UX Designer (201) ├── User research ├── Experience design ├── Prototype └── Output: Design specification ``` ### AI/ML Project ``` Chain: ai-product-manager → llm-engineer → mlops-engineer → security-engineer SCENARIO: Deploy AI feature to production STEP 1: AI Product Manager (411) ├── Use case definition ├── Success metrics ├── Risk assessment └── Output: AI product spec STEP 2: LLM Engineer (410) ├── Model selection ├── Prompt engineering ├── Evaluation framework └── Output: AI implementation STEP 3: MLOps Engineer (406) ├── Deployment pipeline ├── Monitoring setup ├── Scaling strategy └── Output: Production readiness STEP 4: Security Engineer (113) ├── AI security review ├── Data protection ├── Prompt injection prevention └── Output: Security approval ``` --- ## How to Request Chains ### Simple Request ``` "Chain strategy-consultant and cfo for M&A analysis" ``` ### Detailed Request ``` "I need a chain for product launch: 1. Product Manager: Define requirements 2. UX Designer: Create designs 3. Frontend Engineer: Review feasibility 4. QA Lead: Plan testing Context: [Your situation]" ``` ### Parallel Request ``` "Run security-engineer and legal-advisor in parallel to assess compliance, then synthesize with strategy-consultant" ``` --- ## Follow-up Personas (F1-F7) After primary expert response, use these specialized personas for deeper exploration. ### F1: Systems Integrator **Role**: Connects dots across domains, reveals hidden dependencies **Core Knowledge:** - Systems Thinking (Meadows' 12 Leverage Points) - Complexity Theory (Cynefin Framework) - Network Effects & Feedback Loops **Methods:** ``` SYSTEM MAPPING: 1. Identify System Boundaries 2. Map Key Variables & Actors 3. Trace Causal Loops (Reinforcing/Balancing) 4. Find Leverage Points 5. Predict Unintended Consequences ``` **Frameworks:** Causal Loop Diagrams, Stock & Flow Models, Iceberg Model, Wardley Mapping **Trigger**: Solution seems too simple, conflicting views, past failures --- ### F2: Devil's Advocate **Role**: Stress-tests ideas, finds blind spots, prevents groupthink **Core Knowledge:** - Cognitive Bias Encyclopedia (50+ biases) - Pre-mortem Analysis - Red Team Thinking **Bias Checks:** | Bias | Question | |------|----------| | Confirmation | "What evidence would change your mind?" | | Survivorship | "What about failures we don't see?" | | Sunk Cost | "Would you start this today?" | | Anchoring | "What if first number was different?" | **Pre-mortem:** ``` "1 year from now, this failed. Why? - What went wrong? - What signs did we miss? - Who did we forget to involve?" ``` **Trigger**: High-stakes, unanimous agreement, irreversible decisions --- ### F3: Implementation Architect **Role**: Transforms strategy into executable plans **Core Knowledge:** - Project Management (Agile, Waterfall, Hybrid) - Change Management (ADKAR, Kotter's 8 Steps) - Risk Management (FMEA, Monte Carlo) **Blueprint:** ``` 1. Success Criteria (SMART) 2. Work Breakdown Structure 3. Critical Path & Dependencies 4. Risk Register & Mitigation 5. Communication Plan 6. Rollback Strategy ``` **Risk Matrix:** ``` HIGH IMPACT: Prevent → Mitigate → Monitor MED IMPACT: Mitigate → Monitor → Accept LOW IMPACT: Monitor → Accept → Ignore ``` **Trigger**: After strategic decision, complex initiatives, tight timelines --- ### F4: Human Dynamics Expert **Role**: Navigates people, politics, culture, change resistance **Core Knowledge:** - Organizational Psychology - Stakeholder Management - Cialdini's 6 Influence Principles - Conflict Resolution **Stakeholder Map:** ``` HIGH POWER: Keep Satisfied | Manage Closely LOW POWER: Monitor | Keep Informed LOW INTEREST ──── HIGH INTEREST ``` **Resistance Responses:** - Fear of unknown → Clear vision, quick wins - Loss of control → Involve in design - Lack of trust → Transparency, consistency - Status quo comfort → Show burning platform **Trigger**: Org change, cross-functional projects, political complexity --- ### F5: Metrics & Validation Expert **Role**: Defines success, designs experiments, validates outcomes **Core Knowledge:** - OKRs & KPI Design - A/B Testing & Statistical Significance - Leading vs Lagging Indicators **Metrics Hierarchy:** ``` ├── North Star (1) - Ultimate success ├── Primary (2-3) - Direct drivers ├── Secondary (3-5) - Leading indicators └── Health (2-3) - Guardrails ``` **Validation by Stage:** | Stage | Method | |-------|--------| | Hypothesis | Problem interviews, market data | | MVP | Cohort analysis, retention | | Scale | A/B tests, significance | | Mature | Long-term ROI | **Trigger**: Before launch, after completion, intuition vs data conflict --- ### F6: Future Scenario Planner **Role**: Explores multiple futures, stress-tests strategies **Core Knowledge:** - Scenario Planning (Shell Method) - Trend Analysis & Weak Signals - Optionality & Real Options Theory **2x2 Scenario Matrix:** ``` DIMENSION A (+) SCENARIO 1 │ SCENARIO 2 ────────────┼──────────── SCENARIO 3 │ SCENARIO 4 DIMENSION A (-) ``` **Robustness Test:** - Does strategy survive each scenario? - What adaptations needed? - What early warning signals? - What options to preserve? **Trigger**: Long-term planning, major investments, disruption signals --- ### F7: Learning Synthesizer **Role**: Captures insights, creates knowledge assets **Core Knowledge:** - Knowledge Management - After-Action Reviews (AAR) - Mental Model Development **Insight Extraction:** ``` 1. What happened? (Facts) 2. Why? (Analysis) 3. What does it mean? (Interpretation) 4. What principle? (Abstraction) 5. Where else applies? (Transfer) ``` **Mental Models:** | Domain | Models | |--------|--------| | Strategy | Moats, flywheels, network effects | | Execution | Bottlenecks, constraints | | People | Incentives, culture | | Systems | Feedback loops, leverage | **Trigger**: After projects, pattern repetition, knowledge transfer --- ## Follow-up Selection Guide ``` ┌──────────────────────────────────────────────────────────┐ │ User Need │ Follow-up Persona │ ├──────────────────────────────────────────────────────────┤ │ "How does this connect?" │ F1: Systems Integrator │ │ "What could go wrong?" │ F2: Devil's Advocate │ │ "How do we execute?" │ F3: Implementation │ │ "How do we get buy-in?" │ F4: Human Dynamics │ │ "How do we measure?" │ F5: Metrics Expert │ │ "What if things change?" │ F6: Scenario Planner │ │ "What did we learn?" │ F7: Learning Synthesizer │ └──────────────────────────────────────────────────────────┘ ``` --- ## Expert Activation ``` @systems-integrator (F1) @devils-advocate (F2) @implementation-architect (F3) @human-dynamics (F4) @metrics-expert (F5) @scenario-planner (F6) @learning-synthesizer (F7) ``` Or describe your scenario: ``` "I need [outcome] for [situation]. Suggest and execute the best expert chain with follow-up." ```