aiox-core 5.0.2 → 5.0.4

package/pro/squads/squad-creator-pro/docs/MIGRATION-ROADMAP-HYBRIDOPS.md

@@ -0,0 +1,1161 @@
+# ROADMAP DE MIGRAÇÃO: HybridOps Patterns → Squad-Creator
+
+**Data:** 2026-02-10
+**Status:** Planejado
+**Patterns Reference:** `data/hybridops-patterns.md`
+**Última Atualização:** 2026-02-10 (v1.1 - Adicionado Fase 0 + Mind Artifacts)
+
+---
+
+## VISÃO GERAL
+
+Migrar patterns do HybridOps para fortalecer o squad-creator com:
+- **4 Modos do Pedro Valério** (validado com Pedro original)
+- **Veto Conditions** que bloqueiam artefatos ruins
+- **Task Anatomy** padronizado (8 campos)
+- **Heuristics Engine** para validação rigorosa
+- **Axioma 10-Dimensões** para scoring multidimensional
+- **Coherence Validation** para detectar contradições
+- **Integração com workflow** squad-chief → oalanicolas → pedro-valerio
+
+**EXCLUÍDO DO ESCOPO:**
+- ~~Dual-Mode Execution~~ (não aplicável ao squad-creator)
+- ~~ClickUp Integration~~ (não aplicável ao squad-creator)
+
+---
+
+## FASE 0: AGENT UPDATE (ANTES de tudo)
+**Objetivo:** Atualizar pedro-valerio.md com os 4 modos validados
+
+### 0.1 Atualizar Agent File com 4 Modos
+
+**Task:** Implementar estrutura do Squad OPS no agent
+
+**Arquivo a modificar:** `squads/squad-creator-pro/agents/pedro-valerio.md`
+
+**Mudanças:**
+
+```yaml
+# ADICIONAR após PERSONA section
+
+## MODOS DE OPERAÇÃO
+
+modes:
+  engenheiro_processos:
+    name: "Engenheiro de Processos"
+    icon: "🔍"
+    description: "Mapeia processo do fim pro começo, encontra gaps"
+    commands:
+      - "*map-process {processo}" - Mapear processo completo
+      - "*find-gaps {workflow}" - Identificar gaps de tempo
+      - "*identify-owners {processo}" - Descobrir quem faz o quê
+    tools: ["Figma", "Notion", "Google Docs", "Loom"]
+    patterns_used: ["HO-HE-001", "HO-VC-001"]
+    veto_conditions:
+      - "Vision clarity < 0.7"
+      - "Processo sem owner identificado"
+
+  arquiteto_sistemas:
+    name: "Arquiteto de Sistemas"
+    icon: "🏗️"
+    description: "Define estrutura, statuses, campos, permissões"
+    commands:
+      - "*design-structure {sistema}" - Criar estrutura
+      - "*create-statuses {workflow}" - Definir fluxo de status
+      - "*define-fields {entidade}" - Campos personalizados
+    tools: ["ClickUp", "Notion", "Google Drive", "Airtable"]
+    patterns_used: ["HO-TP-001", "HO-EP-001/002/003/004"]
+    veto_conditions:
+      - "Status workflow permite voltar"
+      - "Campos obrigatórios faltando"
+
+  arquiteto_automacao:
+    name: "Arquiteto de Automação"
+    icon: "⚡"
+    description: "Cria regras que bloqueiam erros, conecta sistemas"
+    commands:
+      - "*create-rules {sistema}" - Regras de bloqueio
+      - "*connect-systems {a} {b}" - Integrar sistemas
+      - "*design-triggers {workflow}" - Gatilhos automáticos
+    tools: ["ClickUp Automations", "N8N", "Webhooks", "APIs"]
+    patterns_used: ["HO-HE-003", "HO-VC-003", "HO-QG-001"]
+    veto_conditions:
+      - "Automação sem 5 guardrails"
+      - "Sem manual escape route"
+
+  construtor_templates:
+    name: "Construtor de Templates"
+    icon: "📋"
+    description: "Cria templates replicáveis, testa com pessoa de fora"
+    commands:
+      - "*create-template {tipo}" - Template replicável
+      - "*write-instructions {processo}" - Instruções claras
+      - "*test-leigo {template}" - Teste da filha
+    tools: ["Notion", "ClickUp", "Markdown", "Loom"]
+    patterns_used: ["HO-QG-001", "HO-CV-001", "HO-AX-001"]
+    veto_conditions:
+      - "Template precisa de treinamento"
+      - "Instrução fora do sistema"
+```
+
+**Atualizar Greeting:**
+
+```
+⚙️ **Pedro Valério** - AI Head de OPS
+
+"Tá ligado que processo que permite erro é processo quebrado, né?
+Me passa os insumos que eu construo os artefatos."
+
+**Modos de Operação:**
+🔍 `*eng-` - Engenheiro de Processos (mapear, gaps, owners)
+🏗️ `*arq-` - Arquiteto de Sistemas (estrutura, status, campos)
+⚡ `*auto-` - Arquiteto de Automação (regras, triggers, integrações)
+📋 `*tmpl-` - Construtor de Templates (templates, instruções, teste)
+
+**Comandos de Criação:**
+- `*create-task {name}` - Criar task a partir de insumos
+- `*create-workflow {name}` - Criar workflow multi-fase
+- `*create-agent {name}` - Criar agent a partir de DNA
+
+`*help` para todos os comandos
+```
+
+**Acceptance Criteria:**
+- [ ] 4 modos documentados no agent file
+- [ ] Comandos mapeados por modo
+- [ ] Greeting atualizado
+- [ ] Patterns linkados por modo
+- [ ] Veto conditions por modo
+
+**Esforço:** 4h
+
+---
+
+### 0.2 Copiar Mind Artifacts para Squad-Creator (Self-Contained)
+
+**Task:** Copiar artefatos do HybridOps para dentro do squad-creator
+
+**REGRA:** Squad-creator deve ser 100% self-contained. Nenhuma referência externa.
+
+**Estrutura a criar:**
+```
+squads/squad-creator-pro/
+├── minds/
+│   └── pedro_valerio/
+│       ├── heuristics/
+│       │   ├── PV_BS_001.md       # Future Back-Casting
+│       │   ├── PV_PA_001.md       # Systemic Coherence Scan
+│       │   └── PV_PM_001.md       # Automation Tipping Point
+│       └── artifacts/
+│           ├── META_AXIOMAS.md    # 10 dimensões
+│           └── Assinatura_Linguistica.md  # Voice DNA
+```
+
+**Comandos de cópia:**
+```bash
+# Criar estrutura
+mkdir -p squads/squad-creator-pro/minds/pedro_valerio/{heuristics,artifacts}
+
+# Copiar heuristics
+cp squads/hybrid-ops-squad/minds/pedro_valerio/heuristics/PV_BS_001.md \
+   squads/squad-creator-pro/minds/pedro_valerio/heuristics/
+cp squads/hybrid-ops-squad/minds/pedro_valerio/heuristics/PV_PA_001.md \
+   squads/squad-creator-pro/minds/pedro_valerio/heuristics/
+cp squads/hybrid-ops-squad/minds/pedro_valerio/heuristics/PV_PM_001.md \
+   squads/squad-creator-pro/minds/pedro_valerio/heuristics/
+
+# Copiar artifacts
+cp squads/hybrid-ops-squad/minds/pedro_valerio/artifacts/META_AXIOMAS.md \
+   squads/squad-creator-pro/minds/pedro_valerio/artifacts/
+cp squads/hybrid-ops-squad/minds/pedro_valerio/artifacts/Assinatura_Linguistica.md \
+   squads/squad-creator-pro/minds/pedro_valerio/artifacts/
+```
+
+**Sync com .claude/commands/:**
+```bash
+# Copiar para commands também
+mkdir -p .claude/commands/squad-creator/minds/pedro_valerio/{heuristics,artifacts}
+cp -r squads/squad-creator-pro/minds/pedro_valerio/* \
+   .claude/commands/squad-creator/minds/pedro_valerio/
+```
+
+**Acceptance Criteria:**
+- [ ] 3 heuristics copiados para squad-creator/minds/
+- [ ] 2 artifacts copiados para squad-creator/minds/
+- [ ] Zero referências externas
+- [ ] Squad-creator é 100% self-contained
+- [ ] Sync com .claude/commands/ feito
+
+**Esforço:** 1h
+
+---
+
+### 0.3 Sync com .claude/commands/
+
+**Task:** Sincronizar agent atualizado
+
+**Comando:**
+```bash
+cp squads/squad-creator-pro/agents/pedro-valerio.md .claude/commands/squad-creator/agents/pedro-valerio.md
+```
+
+**Ou usar script existente:**
+```bash
+python squads/squad-creator-pro/scripts/sync-ide-command.py
+```
+
+**Esforço:** 0.5h
+
+---
+
+## FASE 1: FOUNDATION (Semana 1)
+**Objetivo:** Estabelecer base de veto conditions e task anatomy
+
+### 1.1 Implementar Veto Conditions Framework
+
+**Task:** Criar engine de veto conditions em quality gates
+
+**Arquivos a criar:**
+```
+squads/squad-creator-pro/
+├── config/
+│   └── veto-conditions.yaml       # Engine de veto (NEW)
+├── checklists/
+│   └── veto-validation.md         # Checklist de veto (NEW)
+```
+
+**Estrutura `veto-conditions.yaml`:**
+```yaml
+veto_engine:
+  version: "1.0"
+
+  conditions:
+    # Phase 2: Architecture
+    SC_VC_001:
+      name: "Vision Clarity Veto"
+      phase: "architecture"
+      trigger: "squad_vision_clarity < 0.7"
+      action: "VETO - Return to discovery"
+      severity: "BLOCKING"
+      source_pattern: "HO-VC-001"
+
+    # Phase 3: Agent Design
+    SC_VC_002:
+      name: "Agent Coherence Veto"
+      phase: "agent_design"
+      trigger: "agent_behavior_coherence < 0.7"
+      action: "VETO - Reject agent, redesign"
+      severity: "CRITICAL"
+      source_pattern: "HO-VC-002"
+
+    # Phase 4: Workflow Design
+    SC_VC_003:
+      name: "Guardrail Missing Veto"
+      phase: "workflow_design"
+      trigger: "guardrails_present = false"
+      action: "VETO - Define guardrails first"
+      severity: "BLOCKING"
+      required_guardrails:
+        - loop_prevention
+        - idempotency
+        - audit_trail
+        - manual_escape
+        - retry_logic
+      source_pattern: "HO-VC-003"
+
+    # Phase 5: DNA Extraction
+    SC_VC_004:
+      name: "Source Quality Veto"
+      phase: "dna_extraction"
+      trigger: "source_quality_score < 0.6"
+      action: "VETO - Insufficient sources"
+      severity: "BLOCKING"
+
+    # Phase 6: Validation
+    SC_VC_005:
+      name: "Smoke Test Veto"
+      phase: "validation"
+      trigger: "smoke_tests_passed < 3"
+      action: "VETO - Agent behavior invalid"
+      severity: "CRITICAL"
+
+  outcomes:
+    APPROVE: "Proceed to next phase"
+    REVIEW: "Address concerns before proceeding"
+    VETO: "BLOCKED - Cannot proceed until resolved"
+```
+
+**Integração em tasks:**
+- `validate-squad.md` - Adicionar veto checks por fase
+- `create-agent.md` - Adicionar SC_VC_002 check
+- `create-workflow.md` - Adicionar SC_VC_003 check
+
+**Acceptance Criteria:**
+- [ ] `veto-conditions.yaml` criado com 5+ conditions
+- [ ] Veto checks integrados em validate-squad.md
+- [ ] VETO bloqueia progresso até resolução
+- [ ] Log de vetos para auditoria
+
+**Esforço:** 5h
+
+---
+
+### 1.2 Implementar Task Anatomy Validator (8 Campos)
+
+**Task:** Criar validador automático de task anatomy
+
+**Arquivos a criar/modificar:**
+```
+squads/squad-creator-pro/
+├── config/
+│   └── task-anatomy.yaml          # Schema dos 8 campos (NEW)
+├── scripts/
+│   └── task-anatomy-validator.py  # Validador Python (NEW)
+├── checklists/
+│   └── task-anatomy-checklist.md  # UPDATE com enforcement
+```
+
+**Schema `task-anatomy.yaml`:**
+```yaml
+task_anatomy:
+  version: "1.0"
+  source_pattern: "HO-TP-001"
+
+  required_fields:
+    - field: "task_name"
+      type: "string"
+      validation: "starts_with_verb"
+      error: "Task name must start with action verb (e.g., Extract, Validate, Generate)"
+
+    - field: "status"
+      type: "enum"
+      values: ["pending", "in_progress", "review", "completed", "blocked"]
+      error: "Invalid status value"
+
+    - field: "responsible_executor"
+      type: "string"
+      validation: "role_or_agent_ref"
+      error: "Must be role name or @agent reference"
+
+    - field: "execution_type"
+      type: "enum"
+      values: ["Human", "Agent", "Hybrid", "Worker"]
+      error: "Must be one of: Human, Agent, Hybrid, Worker"
+
+    - field: "estimated_time"
+      type: "string"
+      validation: "has_unit"
+      pattern: "\\d+[hmd]|\\d+-\\d+[hmd]"
+      error: "Must include time unit (h, m, d)"
+
+    - field: "input"
+      type: "array"
+      validation: "non_empty"
+      min_items: 1
+      error: "Input array must have at least 1 item"
+
+    - field: "output"
+      type: "array"
+      validation: "non_empty"
+      min_items: 1
+      error: "Output array must have at least 1 item"
+
+    - field: "action_items"
+      type: "array"
+      validation: "non_empty"
+      min_items: 1
+      error: "Action items must have at least 1 step"
+
+  optional_fields:
+    - field: "acceptance_criteria"
+      type: "array"
+      recommended: true
+
+    - field: "tools"
+      type: "array"
+
+    - field: "templates"
+      type: "array"
+
+    - field: "quality_gate"
+      type: "string"
+
+    - field: "handoff"
+      type: "object"
+
+  validation_result:
+    pass: "All 8 required fields present and valid"
+    fail: "VETO - Task anatomy incomplete"
+```
+
+**Acceptance Criteria:**
+- [ ] `task-anatomy.yaml` schema criado
+- [ ] `task-anatomy-validator.py` implementado
+- [ ] Validação integrada em `create-task.md`
+- [ ] Falha em 8-field = VETO automático
+
+**Esforço:** 3h
+
+---
+
+## FASE 2: HEURISTICS ENGINE (Semana 2)
+**Objetivo:** Implementar heuristics de validação com veto power
+
+### 2.1 Implementar SC_HE_001 - Squad Vision Validation
+
+**Task:** Adaptar PV_BS_001 (Future Back-Casting) para squad-creator
+
+**Arquivo:** `squads/squad-creator-pro/config/heuristics.yaml`
+
+```yaml
+heuristics:
+  SC_HE_001:
+    name: "Squad Vision Validation"
+    source_pattern: "HO-HE-001 (PV_BS_001)"
+    phase: "architecture"
+
+    question: "Does this squad architecture enable its stated purpose?"
+
+    weights:
+      squad_purpose_clarity: 0.9
+      agent_coverage: 0.8
+      workflow_completeness: 0.7
+
+    thresholds:
+      high_confidence: 0.8
+      medium_confidence: 0.7
+      veto_threshold: 0.6
+
+    evaluation_points:
+      - "Squad purpose clearly defined?"
+      - "Agents cover all required capabilities?"
+      - "Workflows connect agents properly?"
+      - "Quality gates prevent bad outputs?"
+      - "Handoffs have zero gaps?"
+
+    outcomes:
+      APPROVE: "Squad architecture is sound"
+      REVIEW: "Minor gaps, address before Phase 3"
+      VETO: "Architecture fails to serve purpose"
+```
+
+**Esforço:** 4h
+
+---
+
+### 2.2 Implementar SC_HE_002 - Agent Coherence Scan
+
+**Task:** Adaptar PV_PA_001 (Systemic Coherence) para validar agents
+
+**Adicionar a `heuristics.yaml`:**
+
+```yaml
+  SC_HE_002:
+    name: "Agent Coherence Scan"
+    source_pattern: "HO-HE-002 (PV_PA_001)"
+    phase: "agent_validation"
+
+    purpose: "Validate agent behavior matches documented persona"
+
+    weights:
+      output_consistency: 1.0      # VETO power
+      persona_alignment: 0.9
+      guardrail_compliance: 1.0    # VETO power
+
+    thresholds:
+      coherent: 0.7
+      inconsistent: 0.4
+      incoherent: 0.0
+
+    veto_conditions:
+      - condition: "output_consistency < 0.7"
+        action: "VETO - Agent output varies unexpectedly"
+      - condition: "guardrail_compliance < 1.0"
+        action: "VETO - Agent violates safety guardrails"
+
+    tests:
+      - name: "Consistency Test"
+        method: "Run same input 10x"
+        threshold: "10/10 identical outputs"
+
+      - name: "Persona Alignment Test"
+        method: "Compare output style to voice_dna"
+        threshold: ">90% alignment"
+
+      - name: "Guardrail Compliance Test"
+        method: "Test edge cases against anti_patterns"
+        threshold: "0 violations"
+
+    outcomes:
+      APPROVE: "Agent is coherent and reliable"
+      REVIEW: "Minor inconsistencies, investigate"
+      VETO: "Agent is incoherent, redesign required"
+```
+
+**Esforço:** 4h
+
+---
+
+### 2.3 Implementar SC_HE_003 - Workflow Automation Validation
+
+**Task:** Adaptar PV_PM_001 (Automation Tipping Point) para workflows
+
+**Adicionar a `heuristics.yaml`:**
+
+```yaml
+  SC_HE_003:
+    name: "Workflow Automation Validation"
+    source_pattern: "HO-HE-003 (PV_PM_001)"
+    phase: "workflow_validation"
+
+    purpose: "Ensure workflows have proper automation with guardrails"
+
+    automation_mandate:
+      rule_1: "Task repeated 2+ times → Document and automate"
+      rule_2: "Task repeated 3+ times without automation → Design failure"
+      rule_3: "Any automation MUST have 5 guardrails"
+
+    required_guardrails:
+      - name: "loop_prevention"
+        description: "Max iterations, deduplication"
+      - name: "idempotency"
+        description: "Same input → same output"
+      - name: "audit_trail"
+        description: "Log all state changes"
+      - name: "manual_escape"
+        description: "Human override route"
+      - name: "retry_logic"
+        description: "Graceful failure handling"
+
+    weights:
+      guardrails_present: 1.0    # VETO power
+      checkpoint_coverage: 0.9
+      unidirectional_flow: 0.9
+      handoff_continuity: 0.8
+
+    validation_matrix:
+      - check: "Has checkpoints per phase"
+        required: true
+      - check: "Flow is unidirectional"
+        required: true
+      - check: "Veto conditions defined"
+        required: true
+      - check: "Zero time gaps in handoffs"
+        required: true
+      - check: "All 5 guardrails present"
+        required: true
+        veto_on_fail: true
+
+    outcomes:
+      APPROVE: "Workflow is well-designed"
+      REVIEW: "Missing non-critical elements"
+      VETO: "Missing guardrails or critical flaws"
+```
+
+**Esforço:** 4h
+
+---
+
+## FASE 3: AXIOMA & QUALITY (Semana 3)
+**Objetivo:** Scoring multidimensional e quality gates avançados
+
+### 3.1 Implementar Axioma 10-Dimensões
+
+**Task:** Criar framework de avaliação multidimensional
+
+**Arquivo:** `squads/squad-creator-pro/config/axioma-validator.yaml`
+
+```yaml
+axioma_validator:
+  version: "1.0"
+  source_pattern: "HO-AX-001"
+
+  dimensions:
+    - id: 1
+      name: "Truthfulness"
+      description: "Accuracy and reliability of outputs"
+      weight: 1.0
+      threshold: 7.0
+      veto_power: true
+      squad_creator_application: "Agent outputs are verifiable and consistent"
+
+    - id: 2
+      name: "Coherence"
+      description: "Internal consistency and logic"
+      weight: 0.9
+      threshold: 6.0
+      veto_power: false
+      squad_creator_application: "Agent behavior matches documented persona"
+
+    - id: 3
+      name: "Strategic Alignment"
+      description: "Supports organizational objectives"
+      weight: 0.9
+      threshold: 6.0
+      veto_power: false
+      squad_creator_application: "Squad serves its stated purpose"
+
+    - id: 4
+      name: "Operational Excellence"
+      description: "Efficiency, reliability, documentation"
+      weight: 0.8
+      threshold: 6.0
+      veto_power: false
+      squad_creator_application: "Workflows are efficient and documented"
+
+    - id: 5
+      name: "Innovation Capacity"
+      description: "Ability to improve and adapt"
+      weight: 0.7
+      threshold: 5.0
+      veto_power: false
+      squad_creator_application: "Squad can evolve with new agents/tasks"
+
+    - id: 6
+      name: "Risk Management"
+      description: "Identification and mitigation of risks"
+      weight: 0.8
+      threshold: 6.0
+      veto_power: false
+      squad_creator_application: "Guardrails and veto conditions present"
+
+    - id: 7
+      name: "Resource Optimization"
+      description: "Efficient use of time/money/compute"
+      weight: 0.8
+      threshold: 6.0
+      veto_power: false
+      squad_creator_application: "Worker tasks where possible, Agent only when needed"
+
+    - id: 8
+      name: "Stakeholder Value"
+      description: "Value delivered to all parties"
+      weight: 0.7
+      threshold: 6.0
+      veto_power: false
+      squad_creator_application: "Squad solves real problems"
+
+    - id: 9
+      name: "Sustainability"
+      description: "Long-term viability"
+      weight: 0.7
+      threshold: 6.0
+      veto_power: false
+      squad_creator_application: "No custom code, config-driven"
+
+    - id: 10
+      name: "Adaptability"
+      description: "Capacity for evolution"
+      weight: 0.6
+      threshold: 5.0
+      veto_power: false
+      squad_creator_application: "Squad can be extended easily"
+
+  scoring:
+    formula: "(Σ(score_i × weight_i)) ÷ Σ(weight_i)"
+    pass_threshold: 7.0
+    per_dimension_minimum: 6.0
+    veto_trigger: "truthfulness < 7.0"
+
+  assessment_template: |
+    ## Axioma Assessment: {squad_name}
+
+    | Dimension | Score | Threshold | Status |
+    |-----------|-------|-----------|--------|
+    | Truthfulness | {score_1}/10 | 7.0 | {status_1} |
+    | Coherence | {score_2}/10 | 6.0 | {status_2} |
+    | ... | ... | ... | ... |
+
+    **Overall Score:** {overall}/10
+    **Status:** {APPROVE|REVIEW|VETO}
+```
+
+**Esforço:** 8h
+
+---
+
+### 3.2 Implementar Quality Gates Avançados
+
+**Task:** Criar estrutura de quality gates por tipo
+
+**Arquivo:** `squads/squad-creator-pro/config/quality-gates.yaml`
+
+```yaml
+quality_gates:
+  version: "1.0"
+  source_pattern: "HO-QG-001"
+
+  gate_types:
+    automated:
+      executor: "System/Worker"
+      speed: "<1 second"
+      use_case: "Deterministic checks"
+      examples:
+        - "YAML syntax validation"
+        - "Required fields present"
+        - "Schema compliance"
+
+    hybrid:
+      executor: "System + Human"
+      speed: "1-5 minutes"
+      use_case: "AI prepares, human approves"
+      examples:
+        - "Agent output review"
+        - "DNA extraction validation"
+        - "Template generation check"
+
+    manual:
+      executor: "Human"
+      speed: "1-8 hours"
+      use_case: "Judgment-based decisions"
+      examples:
+        - "Squad acceptance"
+        - "Go-live approval"
+        - "Architecture sign-off"
+
+  gates:
+    # Phase 1: Discovery
+    QG_SC_1_1:
+      name: "Domain Viability Check"
+      phase: 1
+      type: "hybrid"
+      criteria:
+        - "Elite minds exist in domain (≥3)"
+        - "Documented frameworks available"
+        - "Sources are accessible"
+      veto_on_fail: true
+
+    # Phase 2: Architecture
+    QG_SC_2_1:
+      name: "Architecture Validation"
+      phase: 2
+      type: "automated"
+      criteria:
+        - "Squad purpose defined"
+        - "Agent tiers assigned"
+        - "Orchestrator exists"
+      veto_on_fail: true
+
+    # Phase 3: DNA Extraction
+    QG_SC_3_1:
+      name: "DNA Completeness"
+      phase: 3
+      type: "automated"
+      criteria:
+        - "Voice DNA extracted"
+        - "Thinking DNA extracted"
+        - "≥5 axiomas per agent"
+        - "≥5 signature phrases"
+      veto_on_fail: true
+
+    # Phase 4: Agent Creation
+    QG_SC_4_1:
+      name: "Agent Quality Gate"
+      phase: 4
+      type: "hybrid"
+      criteria:
+        - "3/3 smoke tests pass"
+        - "Voice DNA verified with [SOURCE:]"
+        - "Heuristics have WHEN context"
+        - "Handoffs defined"
+      veto_on_fail: true
+
+    # Phase 5: Workflow Integration
+    QG_SC_5_1:
+      name: "Workflow Validation"
+      phase: 5
+      type: "automated"
+      criteria:
+        - "Checkpoints per phase"
+        - "Veto conditions defined"
+        - "Unidirectional flow"
+        - "5 guardrails present"
+      veto_on_fail: true
+
+    # Phase 6: Final Validation
+    QG_SC_6_1:
+      name: "Squad Acceptance"
+      phase: 6
+      type: "manual"
+      criteria:
+        - "Axioma score ≥7.0"
+        - "All previous gates passed"
+        - "Documentation complete"
+      veto_on_fail: true
+```
+
+**Esforço:** 6h
+
+---
+
+## FASE 4: EXECUTOR & COHERENCE (Semana 4)
+**Objetivo:** Executor assignment validation e coherence detection
+
+### 4.1 Implementar Executor Decision Tree
+
+**Task:** Integrar executor-decision-tree.md com validação automática
+
+**Arquivo a modificar:** `squads/squad-creator-pro/data/executor-decision-tree.md`
+
+**Adicionar seção de validação:**
+
+```yaml
+executor_validation:
+  source_patterns:
+    - "HO-EP-001 (Human)"
+    - "HO-EP-002 (Agent)"
+    - "HO-EP-003 (Hybrid)"
+    - "HO-EP-004 (Worker)"
+
+  validation_rules:
+    worker_conditions:
+      - "Task is 100% deterministic"
+      - "No judgment required"
+      - "Can be expressed as if/then rules"
+      - "Examples: template application, field calculation, validation"
+
+    agent_conditions:
+      - "Requires reasoning over data"
+      - "Pattern recognition needed"
+      - "Content generation"
+      - "NOT deterministic (multiple valid outputs)"
+
+    hybrid_conditions:
+      - "Agent prepares, human validates"
+      - "Both judgment types needed"
+      - "High-risk with human gate"
+
+    human_conditions:
+      - "Requires creativity or judgment"
+      - "Stakeholder relationships"
+      - "Strategic decisions"
+      - "Cannot be automated"
+
+  cost_comparison:
+    worker: "$"
+    agent: "$$$$"
+    hybrid: "$$"
+    human: "$$$"
+
+  validation_check:
+    - "Is execution_type aligned with task nature?"
+    - "Could this be Worker instead of Agent? (cost savings)"
+    - "Is Agent overkill for deterministic task?"
+```
+
+**Integração em `create-task.md`:**
+- Adicionar check de executor type vs task nature
+- Sugerir Worker quando Agent é overkill
+- WARN se Agent usado para deterministic task
+
+**Esforço:** 6h
+
+---
+
+### 4.2 Implementar Coherence Validation
+
+**Task:** Criar validador de coerência para agents
+
+**Arquivo:** `squads/squad-creator-pro/scripts/coherence-validator.py`
+
+```python
+"""
+Coherence Validator for Squad-Creator Agents
+Source patterns: HO-CV-001, HO-CV-002
+"""
+
+class CoherenceValidator:
+    COHERENT_THRESHOLD = 0.7
+    INCONSISTENT_THRESHOLD = 0.4
+
+    def validate_agent(self, agent_file: str) -> dict:
+        """
+        Validate agent coherence across multiple dimensions.
+
+        Returns:
+            dict with coherence_score, status, and recommendations
+        """
+        scores = {
+            'capability_alignment': self._check_capability_alignment(),
+            'accountability_clarity': self._check_raci_defined(),
+            'backup_coverage': self._check_backup_exists(),
+            'escalation_paths': self._check_escalation_defined(),
+            'pattern_compliance': self._check_follows_patterns()
+        }
+
+        weights = {
+            'capability_alignment': 0.35,
+            'accountability_clarity': 0.25,
+            'backup_coverage': 0.15,
+            'escalation_paths': 0.10,
+            'pattern_compliance': 0.15
+        }
+
+        weighted_score = sum(
+            scores[k] * weights[k] for k in scores
+        ) / sum(weights.values())
+
+        if weighted_score >= self.COHERENT_THRESHOLD:
+            status = "APPROVE"
+        elif weighted_score >= self.INCONSISTENT_THRESHOLD:
+            status = "REVIEW"
+        else:
+            status = "VETO"
+
+        return {
+            'coherence_score': weighted_score,
+            'status': status,
+            'dimension_scores': scores,
+            'recommendations': self._generate_recommendations(scores)
+        }
+```
+
+**Esforço:** 4h
+
+---
+
+## FASE 5: INTEGRATION & DOCUMENTATION (Semana 5)
+**Objetivo:** Integrar tudo, atualizar squad-chief, e documentar
+
+### 5.1 Integrar Patterns em Workflows
+
+**Task:** Atualizar workflows para usar novos patterns
+
+**Arquivos a modificar:**
+- `workflows/wf-create-squad.yaml` - Adicionar veto checkpoints
+- `workflows/wf-validate-squad.yaml` - Integrar axioma scoring
+- `tasks/validate-squad.md` - Usar quality gates
+
+**Esforço:** 4h
+
+---
+
+### 5.2 Atualizar Squad-Chief com Review Gates
+
+**Task:** Integrar novos quality gates nos comandos de review
+
+**Arquivo a modificar:** `squads/squad-creator-pro/agents/squad-chief.md`
+
+**Adicionar:**
+
+```yaml
+review_commands:
+  review_extraction:
+    uses_patterns:
+      - "HO-HE-002 (Coherence Scan)" - Validar DNA extraído
+      - "HO-CV-001 (Truthfulness)" - Verificar [SOURCE:] tags
+    quality_gates:
+      - "QG_SC_3_1: DNA Completeness"
+    veto_conditions:
+      - "< 15 citações verificáveis"
+      - "< 5 signature phrases"
+
+  review_artifacts:
+    uses_patterns:
+      - "HO-TP-001 (Task Anatomy)" - Validar 8 campos
+      - "HO-HE-003 (Automation)" - Verificar guardrails
+      - "HO-AX-001 (Axioma)" - Scoring final
+    quality_gates:
+      - "QG_SC_4_1: Agent Quality Gate"
+      - "QG_SC_5_1: Workflow Validation"
+    veto_conditions:
+      - "Smoke tests < 3"
+      - "Guardrails missing"
+      - "Axioma score < 7.0"
+
+  routing_to_pedro:
+    modes_available:
+      - "eng-*" → engenheiro_processos
+      - "arq-*" → arquiteto_sistemas
+      - "auto-*" → arquiteto_automacao
+      - "tmpl-*" → construtor_templates
+```
+
+**Acceptance Criteria:**
+- [ ] Review commands usam patterns corretos
+- [ ] Routing para modos do Pedro documentado
+- [ ] Quality gates integrados
+
+**Esforço:** 4h
+
+---
+
+### 5.3 Validar Workflow End-to-End
+
+**Task:** Testar fluxo completo com os novos patterns
+
+**Teste:**
+```
+1. squad-chief recebe demanda
+2. squad-chief delega para oalanicolas
+3. oalanicolas extrai DNA, gera INSUMOS_READY
+4. pedro-valerio (modo correto) constrói artefatos
+5. squad-chief revisa com quality gates
+6. Veto conditions funcionam (testar falha)
+7. Axioma scoring calcula corretamente
+```
+
+**Acceptance Criteria:**
+- [ ] Fluxo completo funciona
+- [ ] Handoff template INSUMOS_READY validado
+- [ ] Veto conditions bloqueiam quando devem
+- [ ] Routing para modos correto
+
+**Esforço:** 4h
+
+---
+
+### 5.4 Criar Pattern Library Consolidado
+
+**Task:** Documentar todos os patterns disponíveis
+
+**Arquivo:** `squads/squad-creator-pro/docs/PATTERN-LIBRARY.md`
+
+Conteúdo:
+- Lista de todos patterns implementados
+- Quando usar cada pattern
+- Exemplos de aplicação
+- Cross-reference com HybridOps
+- Mapeamento modo → pattern
+
+**Esforço:** 3h
+
+---
+
+### 5.5 Atualizar Documentação
+
+**Task:** Atualizar docs com novos patterns
+
+**Arquivos a modificar:**
+- `docs/CONCEPTS.md` - Adicionar Veto Conditions, Axioma, 4 Modos
+- `docs/FAQ.md` - Perguntas sobre quality gates e modos
+- `README.md` - Mencionar heuristics engine e estrutura Squad OPS
+
+**Esforço:** 3h
+
+---
+
+## CRONOGRAMA CONSOLIDADO (REVISADO v1.1)
+
+| Semana | Fase | Deliverables | Esforço |
+|--------|------|--------------|---------|
+| **0** | Agent Update | pedro-valerio.md + 4 modos + Mind COPY + Sync | **7h** |
+| **1** | Foundation | Veto Conditions + Task Anatomy | 8h |
+| **2** | Heuristics | SC_HE_001, SC_HE_002, SC_HE_003 | 12h |
+| **3** | Quality | Axioma 10-Dim + Quality Gates | 14h |
+| **4** | Executor | Executor Validation + Coherence | 10h |
+| **5** | Integration | Workflows + squad-chief + Docs + Test E2E | **18h** |
+
+**Total:** ~69h (6 semanas, ~11.5h/semana)
+
+**REGRA:** Squad-creator é 100% self-contained. Mind artifacts são COPIADOS, não referenciados.
+
+**Breakdown Fase 0:**
+- 0.1 Agent file com 4 modos: 4h
+- 0.2 Copiar mind artifacts: 1h
+- 0.3 Sync com .claude/commands/: 0.5h
+- 0.4 Testar ativação: 1.5h
+
+### Dependências
+
+```
+Fase 0 ──┬──> Fase 1 ──> Fase 2 ──> Fase 3
+         │
+         └──> Fase 4 ──────────────> Fase 5
+```
+
+- Fase 0 é PRÉ-REQUISITO para todas as outras
+- Fases 1-4 podem ter paralelismo parcial
+- Fase 5 depende de TODAS as anteriores
+
+---
+
+## CRITÉRIOS DE SUCESSO
+
+### Métricas de Qualidade
+
+| Métrica | Antes | Meta |
+|---------|-------|------|
+| **Pedro Valério tem 4 modos** | 0 | 4 |
+| **Comandos mapeados por modo** | 0 | 12+ |
+| Agents com veto check | 0% | 100% |
+| Tasks com 8 campos | ~60% | 100% |
+| Workflows com guardrails | ~40% | 100% |
+| Squads com axioma score | 0% | 100% |
+| **Mind artifacts copiados** | 0 | 5 |
+| **Squad-chief usa quality gates** | 0 | 100% |
+
+### Definição de Done
+
+**Fase 0 (CRÍTICO):**
+- [ ] pedro-valerio.md tem 4 modos documentados
+- [ ] Comandos mapeados: `*eng-*`, `*arq-*`, `*auto-*`, `*tmpl-*`
+- [ ] Greeting atualizado com modos
+- [ ] Mind artifacts COPIADOS para minds/pedro_valerio/
+- [ ] Sync com .claude/commands/ feito
+
+**Fases 1-4:**
+- [ ] Veto conditions bloqueiam artefatos ruins
+- [ ] Task anatomy é enforced automaticamente
+- [ ] Heuristics engine valida agents/workflows
+- [ ] Axioma score é calculado para todo squad
+- [ ] Quality gates têm 3 tipos (auto/hybrid/manual)
+- [ ] Coherence validation detecta contradições
+
+**Fase 5:**
+- [ ] Squad-chief usa quality gates em reviews
+- [ ] Workflow E2E testado e funcionando
+- [ ] Handoff INSUMOS_READY validado
+- [ ] Pattern library está documentado
+- [ ] Todos os patterns têm source_pattern linkado
+
+---
+
+## RISCOS E MITIGAÇÕES
+
+| Risco | Probabilidade | Impacto | Mitigação |
+|-------|---------------|---------|-----------|
+| Over-engineering | Média | Alto | Começar com P1 patterns apenas |
+| Veto muito agressivo | Média | Médio | Tunning iterativo de thresholds |
+| Adoção lenta | Baixa | Médio | Documentação clara + exemplos |
+| Incompatibilidade | Baixa | Alto | Testar em squad existente primeiro |
+
+---
+
+## PRÓXIMOS PASSOS IMEDIATOS
+
+### Semana 0 (FAZER AGORA)
+
+1. **Atualizar `agents/pedro-valerio.md`** com 4 modos - 4h
+2. **Copiar mind artifacts** de HybridOps para squad-creator - 1h
+   ```bash
+   mkdir -p squads/squad-creator-pro/minds/pedro_valerio/{heuristics,artifacts}
+   cp squads/hybrid-ops-squad/minds/pedro_valerio/heuristics/*.md \
+      squads/squad-creator-pro/minds/pedro_valerio/heuristics/
+   cp squads/hybrid-ops-squad/minds/pedro_valerio/artifacts/*.md \
+      squads/squad-creator-pro/minds/pedro_valerio/artifacts/
+   ```
+3. **Atualizar greeting** com novos comandos - 0.5h
+4. **Sync com `.claude/commands/`** - 0.5h
+5. **Testar ativação** do Pedro com novos modos - 1h
+
+### Semana 1 (Após Fase 0)
+
+1. **Criar `config/veto-conditions.yaml`** - 2h
+2. **Criar `config/task-anatomy.yaml`** - 1h
+3. **Atualizar `validate-squad.md`** com veto checks - 2h
+4. **Testar em squad existente** (copy ou hormozi) - 3h
+
+---
+
+## CHANGELOG
+
+| Versão | Data | Mudanças |
+|--------|------|----------|
+| v1.2 | 2026-02-10 | Self-contained: COPIAR artifacts em vez de referenciar |
+| v1.1 | 2026-02-10 | Adicionado Fase 0, Mind Artifacts, Squad-chief integration |
+| v1.0 | 2026-02-10 | Versão inicial |
+
+---
+
+*Roadmap v1.2 - 2026-02-10*