specweave 0.1.0

package/src/skills/role-orchestrator/test-cases/test-1-simple-product.yaml

@@ -0,0 +1,98 @@
+---
+name: "Simple Product Development Orchestration"
+description: "Tests role-orchestrator's ability to coordinate agents for simple product development"
+priority: P1
+test_type: integration
+
+input:
+  prompt: "Build a simple todo application with React frontend and Node.js backend"
+  project_context:
+    type: "greenfield"
+    has_specs: false
+    has_architecture: false
+
+expected_behavior:
+  - orchestrator_activates: true
+  - complexity_analysis:
+      scope: "full-product"
+      agents_needed: ["pm-agent", "architect-agent", "nodejs-backend", "frontend-agent", "qa-lead-agent", "devops-agent"]
+      pattern: "sequential_with_gates"
+
+  - execution_phases:
+      - phase: 1
+        name: "Requirements"
+        agent: "pm-agent"
+        output_files:
+          - "specifications/modules/todo-app/overview.md"
+        gate: "user_approval"
+
+      - phase: 2
+        name: "Architecture"
+        agent: "architect-agent"
+        input_from: "pm-agent"
+        output_files:
+          - ".specweave/docs/architecture/todo-app-design.md"
+        gate: "tech_review"
+
+      - phase: 3
+        name: "Feature Planning"
+        agent: "increment-planner"
+        output_files:
+          - "features/001-todo-app/spec.md"
+          - "features/001-todo-app/plan.md"
+          - "features/001-todo-app/tasks.md"
+
+      - phase: 4
+        name: "Implementation"
+        agents: ["nodejs-backend", "frontend-agent"]
+        pattern: "parallel"
+        output_files:
+          - "src/backend/todo-service.ts"
+          - "src/frontend/TodoApp.tsx"
+
+      - phase: 5
+        name: "Testing"
+        agent: "qa-lead-agent"
+        output_files:
+          - "features/001-todo-app/tests.md"
+          - "tests/e2e/todo-app.spec.ts"
+        gate: "test_coverage_80%"
+
+      - phase: 6
+        name: "Deployment"
+        agent: "devops-agent"
+        output_files:
+          - ".github/workflows/deploy.yml"
+        gate: "deployment_approval"
+
+expected_output:
+  success: true
+  phases_completed: 6
+  gates_passed: 4
+  files_created:
+    - "specifications/modules/todo-app/overview.md"
+    - ".specweave/docs/architecture/todo-app-design.md"
+    - "features/001-todo-app/spec.md"
+    - "src/backend/todo-service.ts"
+    - "src/frontend/TodoApp.tsx"
+    - "tests/e2e/todo-app.spec.ts"
+
+  metrics:
+    total_agents_invoked: 6
+    parallel_executions: 1
+    sequential_phases: 5
+    quality_gates_passed: 4
+
+validation:
+  - "All 6 phases completed in correct order"
+  - "Parallel execution of backend and frontend worked"
+  - "All quality gates evaluated correctly"
+  - "Shared context maintained across agents"
+  - "Final output includes all expected files"
+
+success_criteria:
+  - orchestration_complete: true
+  - no_phase_failures: true
+  - all_gates_passed: true
+  - correct_agent_sequence: true
+---

package/src/skills/role-orchestrator/test-cases/test-2-quality-gate-failure.yaml

@@ -0,0 +1,73 @@
+---
+name: "Quality Gate Failure Handling"
+description: "Tests role-orchestrator's ability to handle quality gate failures and coordinate remediation"
+priority: P1
+test_type: integration
+
+input:
+  prompt: "Implement payment processing feature"
+  project_context:
+    type: "brownfield"
+    existing_code: "src/payments/"
+    test_coverage: "45%"  # Below 80% threshold
+
+simulated_scenario:
+  - phase: "implementation"
+    status: "completed"
+    test_coverage_actual: "67%"  # Below target of 80%
+
+  - phase: "qa_gate"
+    gate_type: "test_coverage"
+    target: "80%"
+    actual: "67%"
+    status: "failed"
+
+expected_behavior:
+  - gate_failure_detected: true
+  - orchestrator_response:
+      - pause_deployment: true
+      - invoke_agent: "qa-lead-agent"
+      - task: "Analyze test coverage gaps and create additional tests"
+
+  - qa_lead_analysis:
+      files_examined:
+        - "src/payments/payment-service.ts"
+        - "tests/unit/payment-service.test.ts"
+      gaps_identified:
+        - "Error handling not tested (15% of code)"
+        - "Edge cases missing (10%)"
+        - "Integration tests needed (8%)"
+
+  - orchestrator_next_steps:
+      - create_work_items:
+          - "work/issues/001-improve-test-coverage/"
+      - assign_to: "nodejs-backend"
+      - task: "Implement additional test cases"
+
+  - retry_after_fixes:
+      test_coverage_new: "84%"
+      gate_status: "passed"
+      proceed_to_next_phase: true
+
+expected_output:
+  gate_failure_handled: true
+  remediation_plan_created: true
+  work_items_created:
+    - "work/issues/001-improve-test-coverage/issue.md"
+  final_gate_status: "passed"
+  deployment_proceeded: true
+
+validation:
+  - "Gate failure detected immediately"
+  - "Orchestrator paused next phases"
+  - "QA Lead agent invoked for analysis"
+  - "Work items created for fixes"
+  - "After fixes, gate re-evaluated"
+  - "Gate passed, deployment proceeded"
+
+success_criteria:
+  - failure_detection: "immediate"
+  - remediation_invoked: true
+  - gate_retry_successful: true
+  - no_deployment_before_fix: true
+---

package/src/skills/role-orchestrator/test-cases/test-3-security-workflow.yaml

@@ -0,0 +1,121 @@
+---
+name: "Security Hardening Multi-Agent Workflow"
+description: "Tests role-orchestrator's ability to coordinate security review with multiple agents"
+priority: P1
+test_type: integration
+
+input:
+  prompt: "Perform security audit and harden our authentication system"
+  project_context:
+    type: "brownfield"
+    existing_system: "src/auth/"
+    security_critical: true
+    compliance_required: ["OWASP Top 10", "OAuth 2.0"]
+
+expected_behavior:
+  - orchestrator_pattern: "assess_fix_verify"
+
+  - execution_phases:
+      - phase: 1
+        name: "Security Assessment"
+        agent: "security-agent"
+        task: "Threat modeling and security audit"
+        output_files:
+          - ".specweave/docs/architecture/security/threat-model.md"
+          - ".specweave/docs/architecture/security/vulnerabilities.md"
+        findings:
+          critical:
+            - "SQL injection vulnerability in login endpoint"
+            - "Missing rate limiting on password reset"
+          high:
+            - "Weak password policy"
+            - "Session tokens not rotated"
+          medium:
+            - "HTTPS not enforced on all routes"
+
+      - phase: 2
+        name: "Prioritization"
+        agent: "tech-lead-agent"
+        input_from: "security-agent"
+        task: "Create prioritized fix plan"
+        output_files:
+          - "work/issues/001-critical-security-fixes/issue.md"
+          - "work/issues/002-high-priority-security/issue.md"
+
+      - phase: 3
+        name: "Implementation"
+        agents: ["nodejs-backend"]  # Since auth is Node.js
+        task: "Implement security fixes"
+        fixes_implemented:
+          - "Parameterized queries for SQL injection"
+          - "Rate limiting middleware (express-rate-limit)"
+          - "Enhanced password policy (zxcvbn)"
+          - "Session token rotation"
+          - "HTTPS enforcement"
+        output_files:
+          - "src/auth/auth-service.ts"  # Updated
+          - "src/middleware/rate-limiter.ts"  # New
+          - "src/auth/password-policy.ts"  # New
+
+      - phase: 4
+        name: "Security Testing"
+        agent: "security-agent"
+        task: "Penetration testing"
+        tests_performed:
+          - "SQL injection attempts"
+          - "Brute force attacks"
+          - "Session hijacking"
+          - "CSRF attacks"
+        gate: "no_critical_vulnerabilities"
+
+      - phase: 5
+        name: "Regression Testing"
+        agent: "qa-lead-agent"
+        task: "Ensure no functionality broken"
+        output_files:
+          - "tests/security/auth-security.spec.ts"
+        gate: "all_tests_pass"
+
+      - phase: 6
+        name: "Deployment"
+        agent: "devops-agent"
+        task: "Zero-downtime deployment with rollback plan"
+        strategy: "blue-green"
+        gate: "deployment_approval"
+
+expected_output:
+  success: true
+  vulnerabilities_fixed:
+    critical: 2
+    high: 2
+    medium: 1
+
+  security_gates_passed:
+    - "no_critical_vulnerabilities"
+    - "penetration_tests_passed"
+    - "regression_tests_passed"
+
+  files_created:
+    - ".specweave/docs/architecture/security/threat-model.md"
+    - "work/issues/001-critical-security-fixes/issue.md"
+    - "src/middleware/rate-limiter.ts"
+    - "tests/security/auth-security.spec.ts"
+
+  deployment_status: "successful"
+  rollback_plan: "documented"
+
+validation:
+  - "Security agent performed comprehensive audit"
+  - "Critical vulnerabilities prioritized first"
+  - "All fixes implemented by appropriate agent"
+  - "Penetration testing validated fixes"
+  - "No regressions introduced"
+  - "Deployment successful with rollback plan"
+
+success_criteria:
+  - all_critical_vulns_fixed: true
+  - penetration_tests_passed: true
+  - no_regressions: true
+  - deployment_successful: true
+  - compliance_met: ["OWASP Top 10", "OAuth 2.0"]
+---

package/src/skills/role-orchestrator/test-cases/test-4-parallel-execution.yaml

@@ -0,0 +1,145 @@
+---
+name: "Parallel Agent Execution with Sync Points"
+description: "Tests role-orchestrator's ability to execute agents in parallel and synchronize at key points"
+priority: P2
+test_type: integration
+
+input:
+  prompt: "Build a real-time chat application with message history and notifications"
+  project_context:
+    type: "greenfield"
+    complexity: "high"
+    requires_parallel_work: true
+
+expected_behavior:
+  - orchestrator_pattern: "parallel_with_sync_points"
+
+  - execution_phases:
+      - phase: 1
+        name: "Architecture Design"
+        agent: "architect-agent"
+        output:
+          components:
+            - "WebSocket Server (Backend)"
+            - "React Chat UI (Frontend)"
+            - "Notification Service (Backend)"
+            - "Message Storage (Database)"
+
+      - phase: 2
+        name: "Parallel Implementation"
+        pattern: "parallel"
+        agents:
+          - agent: "nodejs-backend"
+            task: "Implement WebSocket server and message API"
+            files:
+              - "src/backend/websocket-server.ts"
+              - "src/backend/message-service.ts"
+              - "src/backend/message-repository.ts"
+
+          - agent: "frontend-agent"
+            task: "Implement chat UI components"
+            files:
+              - "src/frontend/components/ChatWindow.tsx"
+              - "src/frontend/components/MessageList.tsx"
+              - "src/frontend/hooks/useWebSocket.ts"
+
+          - agent: "nodejs-backend"  # Different instance/task
+            task: "Implement notification service"
+            files:
+              - "src/backend/notification-service.ts"
+              - "src/backend/notification-queue.ts"
+
+        sync_point: "api_contract"
+        shared_artifacts:
+          - "features/002-chat-app/api-contracts.md"  # All agents reference this
+
+      - phase: 3
+        name: "Integration Point"
+        agent: "tech-lead-agent"
+        task: "Verify all components integrate correctly"
+        validation:
+          - "WebSocket API matches frontend expectations"
+          - "Notification service integrates with message service"
+          - "Database schema supports all features"
+
+      - phase: 4
+        name: "Parallel Testing"
+        pattern: "parallel"
+        agents:
+          - agent: "qa-lead-agent"
+            task: "E2E tests for chat functionality"
+            files:
+              - "tests/e2e/chat-flow.spec.ts"
+
+          - agent: "qa-lead-agent"  # Different task
+            task: "Integration tests for notification service"
+            files:
+              - "tests/integration/notification.test.ts"
+
+          - agent: "performance-agent"
+            task: "Load testing for WebSocket server"
+            files:
+              - "tests/performance/websocket-load.test.ts"
+
+        sync_point: "all_tests_pass"
+
+      - phase: 5
+        name: "Deployment"
+        agent: "devops-agent"
+        task: "Deploy with infrastructure for WebSocket support"
+
+expected_output:
+  success: true
+  parallel_executions: 2  # Phase 2 and Phase 4
+
+  phase_2_timing:
+    backend_websocket: "started: 0s, completed: 45s"
+    frontend_ui: "started: 0s, completed: 40s"  # Finished before backend
+    backend_notifications: "started: 0s, completed: 35s"
+    all_completed: "45s"  # Waited for slowest
+    time_saved: "80s"  # vs sequential (45 + 40 + 35 = 120s)
+
+  sync_point_1:
+    artifact: "features/002-chat-app/api-contracts.md"
+    all_agents_referenced: true
+    conflicts_detected: false
+
+  phase_4_timing:
+    e2e_tests: "started: 0s, completed: 30s"
+    integration_tests: "started: 0s, completed: 20s"
+    load_tests: "started: 0s, completed: 40s"
+    all_completed: "40s"
+    all_tests_passed: true
+
+  files_created:
+    backend:
+      - "src/backend/websocket-server.ts"
+      - "src/backend/message-service.ts"
+      - "src/backend/notification-service.ts"
+    frontend:
+      - "src/frontend/components/ChatWindow.tsx"
+      - "src/frontend/hooks/useWebSocket.ts"
+    tests:
+      - "tests/e2e/chat-flow.spec.ts"
+      - "tests/integration/notification.test.ts"
+      - "tests/performance/websocket-load.test.ts"
+
+  total_time: "~150s"
+  time_vs_sequential: "saved ~90s (60% faster)"
+
+validation:
+  - "Agents started in parallel successfully"
+  - "No file conflicts between parallel agents"
+  - "Shared API contract respected by all"
+  - "Sync points enforced correctly"
+  - "All agents waited at sync points"
+  - "Integration validated before proceeding"
+  - "Parallel execution significantly faster"
+
+success_criteria:
+  - parallel_execution_successful: true
+  - no_conflicts: true
+  - sync_points_enforced: true
+  - time_savings: ">50%"
+  - all_tests_passed: true
+---

package/src/skills/role-orchestrator/test-cases/test-5-feedback-loops.yaml

@@ -0,0 +1,149 @@
+---
+name: "Feedback Loops - PM Agent Auto-Refinement"
+description: "Test automatic quality improvement through iterative refinement with feedback"
+skill: role-orchestrator
+priority: P1
+estimated_time: "5-7 minutes"
+
+input:
+  user_request: "Create requirements for user authentication feature"
+
+  config:
+    feedback_loops:
+      enabled: true
+      max_retries: 3
+      stop_on_improvement: true
+      quality_threshold: 0.80
+
+    validation:
+      use_llm_judge: true
+      combine_with_rules: true
+      judge_weight: 0.5
+
+  agent: "pm-agent"
+  task: "Define requirements for user authentication"
+
+expected_output:
+  # Attempt 1: Initial generation (below threshold)
+  attempt_1:
+    output: "Initial requirements spec"
+    validation:
+      rule_based_score: 1.00  # All structural rules pass
+      llm_judge_score: 0.60-0.70
+      combined_score: 0.70-0.85
+      threshold: 0.80
+
+    issues_found:
+      - severity: "major"
+        description: "Acceptance criteria not testable"
+        location: "Success Criteria section"
+
+      - severity: "major"
+        description: "Security requirements too vague"
+        location: "Security section"
+
+      - severity: "minor"
+        description: "No rate limiting specified"
+        location: "Requirements"
+
+    below_threshold: true
+    triggers_refinement: true
+
+  # Attempt 2: Refinement with feedback
+  attempt_2:
+    feedback_provided: |
+      Issues from previous attempt:
+      • MAJOR: Acceptance criteria not testable
+        Location: Success Criteria section
+        Suggestion: Use measurable metrics (e.g., "Login completes in <2s")
+
+      • MAJOR: Security requirements too vague
+        Location: Security section
+        Suggestion: Specify encryption, hashing, rate limiting
+
+      • MINOR: No rate limiting specified
+        Location: Requirements
+        Suggestion: Add brute-force protection (5 attempts, 15min lockout)
+
+    output: "Improved requirements spec with feedback incorporated"
+    validation:
+      rule_based_score: 1.00
+      llm_judge_score: 0.80-0.90
+      combined_score: 0.85-0.95
+      threshold: 0.80
+
+    improvements:
+      - "Acceptance criteria now measurable with metrics"
+      - "Security requirements detailed (bcrypt, JWT, rate limiting)"
+      - "Rate limiting specified (5 attempts, 15min lockout)"
+
+    above_threshold: true
+    stops_refinement: true
+
+  final_result:
+    total_attempts: 2
+    final_score: 0.85-0.95
+    met_threshold: true
+    output_quality: "good"
+
+    remaining_issues:
+      count: 0-1
+      severity: "minor"
+      examples:
+        - "Performance requirements could be more specific (p95/p99 targets)"
+
+validation:
+  # Attempt 1 validation
+  - "Initial attempt produces output"
+  - "Rule-based validation passes (1.00)"
+  - "LLM judge score < 0.80"
+  - "Combined score may or may not be < 0.80 depending on weights"
+  - "At least 2 major issues found"
+  - "Triggers refinement"
+
+  # Attempt 2 validation
+  - "Feedback generated from issues"
+  - "PM agent receives feedback prompt"
+  - "Second attempt incorporates feedback"
+  - "LLM judge score improves (>= 0.80)"
+  - "Combined score >= 0.80"
+  - "Stops after reaching threshold"
+
+  # Overall validation
+  - "Total attempts <= 3"
+  - "Final score >= 0.80"
+  - "Quality improvement measurable"
+  - "Progress shown to user"
+
+success_criteria:
+  - "Feedback loop triggers on low quality"
+  - "Agent successfully incorporates feedback"
+  - "Quality score improves significantly"
+  - "Stops when threshold met (doesn't waste retries)"
+  - "User sees progress updates"
+  - "Final output meets quality standards"
+
+workflow_steps:
+  1. "PM agent generates initial requirements"
+  2. "Validate with rule-based + LLM-as-judge"
+  3. "Detect score below threshold (0.80)"
+  4. "Generate feedback from issues"
+  5. "Reinvoke PM agent with feedback"
+  6. "Validate improved output"
+  7. "Detect score above threshold"
+  8. "Stop refinement, return final output"
+
+edge_cases:
+  max_retries_reached:
+    description: "If quality doesn't improve after 3 attempts"
+    expected: "Return best attempt with warning"
+
+  stop_on_improvement:
+    description: "If score improves but doesn't reach threshold"
+    config: "stop_on_improvement: true"
+    expected: "Stop early if improvement detected"
+
+  first_attempt_passes:
+    description: "If initial attempt >= threshold"
+    expected: "No refinement needed, accept immediately"
+---