specweave 0.1.8 → 0.3.0

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

@@ -1,145 +0,0 @@
----
-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

@@ -1,149 +0,0 @@
----
-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"
----

package/src/skills/skill-creator/test-cases/test-1-placeholder.yaml

@@ -1,12 +0,0 @@
----
-name: "Placeholder Test"
-description: "TODO: Add actual test case for this skill"
-input:
-  prompt: "Test prompt"
-expected_output:
-  type: "validation"
-validation:
-  - "Placeholder validation"
-success_criteria:
-  - "Test passes"
----

package/src/skills/skill-creator/test-cases/test-2-placeholder.yaml

@@ -1,12 +0,0 @@
----
-name: "Placeholder Test"
-description: "TODO: Add actual test case for this skill"
-input:
-  prompt: "Test prompt"
-expected_output:
-  type: "validation"
-validation:
-  - "Placeholder validation"
-success_criteria:
-  - "Test passes"
----

package/src/skills/skill-creator/test-cases/test-3-placeholder.yaml

@@ -1,12 +0,0 @@
----
-name: "Placeholder Test"
-description: "TODO: Add actual test case for this skill"
-input:
-  prompt: "Test prompt"
-expected_output:
-  type: "validation"
-validation:
-  - "Placeholder validation"
-success_criteria:
-  - "Test passes"
----

package/src/skills/skill-router/test-cases/test-1-basic-routing.yaml

@@ -1,33 +0,0 @@
----
-name: "Basic Request Routing"
-description: "Tests if skill-router correctly routes user request to appropriate skill"
-input:
-  prompt: "Plan a new feature for user authentication"
-  context:
-    available_skills:
-      - "increment-planner"
-      - "spec-author"
-      - "developer"
-  user_intent: "plan feature"
-expected_output:
-  type: "skill_routing"
-  routes_to: "increment-planner"
-  confidence: "high"
-  reason: "User request contains 'plan' and 'feature' keywords"
-  alternatives: []
-  actions:
-    - "Activate increment-planner skill"
-    - "Pass user request to increment-planner"
-    - "Log routing decision"
-validation:
-  - "Routes to correct skill (increment-planner)"
-  - "Confidence level is high (>0.9)"
-  - "No ambiguity detected"
-  - "Logs routing decision"
-  - "Activates target skill"
-success_criteria:
-  - "Correct skill selected"
-  - ">90% routing accuracy"
-  - "Fast routing (<100ms)"
-  - "Logging works"
----

package/src/skills/skill-router/test-cases/test-2-ambiguous-request.yaml

@@ -1,42 +0,0 @@
----
-name: "Ambiguous Request Handling"
-description: "Tests how skill-router handles requests that could match multiple skills"
-input:
-  prompt: "Create documentation for the API"
-  context:
-    available_skills:
-      - "spec-author"      # Could create API specification
-      - "architect"        # Could document API architecture
-      - "docs-updater"     # Could update API docs
-  user_intent: "create documentation"
-expected_output:
-  type: "clarification_required"
-  routes_to: null
-  confidence: "low"
-  reason: "Multiple skills match: spec-author, architect, docs-updater"
-  alternatives:
-    - skill: "spec-author"
-      match_score: 0.65
-      reason: "Can create API specifications"
-    - skill: "architect"
-      match_score: 0.60
-      reason: "Can document API architecture"
-    - skill: "docs-updater"
-      match_score: 0.55
-      reason: "Can update API documentation"
-  actions:
-    - "Ask user for clarification"
-    - "Present alternatives with descriptions"
-    - "Wait for user selection"
-validation:
-  - "Detects ambiguity (confidence <0.7)"
-  - "Lists all matching skills"
-  - "Provides match scores"
-  - "Asks user to clarify"
-  - "Does not make incorrect assumption"
-success_criteria:
-  - "Ambiguity detected"
-  - "User presented with options"
-  - "No forced routing when uncertain"
-  - "Helpful context for each option"
----

package/src/skills/skill-router/test-cases/test-3-nested-orchestration.yaml

@@ -1,50 +0,0 @@
----
-name: "Nested Skill Orchestration"
-description: "Tests skill-router's ability to handle one skill calling another skill"
-input:
-  prompt: "Implement the user authentication feature"
-  context:
-    available_skills:
-      - "developer"
-      - "context-loader"
-      - "spec-author"
-    current_skill: "developer"
-  workflow:
-    - step: 1
-      skill: "developer"
-      action: "Needs to load context from specification"
-      calls: "context-loader"
-    - step: 2
-      skill: "context-loader"
-      action: "Load authentication spec"
-      returns_to: "developer"
-    - step: 3
-      skill: "developer"
-      action: "Implement based on loaded context"
-expected_output:
-  type: "nested_routing"
-  primary_skill: "developer"
-  sub_skills:
-    - "context-loader"
-  orchestration:
-    - "developer activates"
-    - "developer calls context-loader"
-    - "context-loader loads spec"
-    - "context returns to developer"
-    - "developer implements"
-  contains:
-    - "Orchestrating skills: developer -> context-loader"
-    - "Context loaded successfully"
-    - "Returning to developer"
-validation:
-  - "Primary skill (developer) activated correctly"
-  - "Sub-skill (context-loader) called when needed"
-  - "Context passed between skills"
-  - "Control returns to primary skill"
-  - "Workflow completes successfully"
-success_criteria:
-  - "Nested skill calls work"
-  - "Context preserved across skills"
-  - "No infinite loops"
-  - "Proper error handling in chain"
----

package/src/skills/spec-driven-brainstorming/test-cases/TC-001-simple-idea-to-design.yaml

@@ -1,148 +0,0 @@
----
-name: "Simple Idea to Validated Design"
-description: "Transform a simple rough idea into a validated design ready for increment creation"
-skill: spec-driven-brainstorming
-priority: P1
-estimated_time: "10-15 minutes"
-
-input:
-  prompt: "I want to add user authentication to my app"
-  context:
-    existing_project: true
-    tech_stack: "Next.js 14, TypeScript"
-    existing_files:
-      - "package.json"
-      - "src/app/page.tsx"
-      - "src/components/"
-    database: "PostgreSQL"
-
-expected_output:
-  phase_1_understanding:
-    questions_asked:
-      - "What's your primary authentication goal?" # user login, API keys, SSO
-      - "Who are the users?" # internal team, public, B2B customers
-      - "Any compliance requirements?" # GDPR, HIPAA, SOC2
-
-    gathered_requirements:
-      purpose: "Secure user login for public SaaS application"
-      constraints:
-        - "Must work with existing PostgreSQL database"
-        - "Budget-friendly (prefer open-source)"
-      success_criteria:
-        - "Users can register and login securely"
-        - "Session management across devices"
-        - "Password reset functionality"
-
-  phase_2_tech_stack:
-    detected_stack:
-      framework: "Next.js 14"
-      language: "TypeScript"
-      database: "PostgreSQL"
-
-    skills_activated:
-      - nextjs
-      - nodejs-backend
-
-    recommendation: "Use NextAuth.js v5 (Auth.js) - native Next.js integration"
-
-  phase_3_exploration:
-    approaches_proposed: 3
-
-    approach_1:
-      name: "NextAuth.js with Credentials Provider"
-      architecture: "Session-based with JWT"
-      trade_offs:
-        pros:
-          - "Simple setup, well-documented"
-          - "Built-in session management"
-        cons:
-          - "Basic features only"
-          - "Custom UI needed"
-      complexity: "Low"
-
-    approach_2:
-      name: "Clerk Integration"
-      architecture: "Third-party auth SaaS"
-      trade_offs:
-        pros:
-          - "Pre-built UI components"
-          - "Advanced features (MFA, social login)"
-        cons:
-          - "Monthly cost ($25+)"
-          - "Vendor lock-in"
-      complexity: "Very Low"
-
-    approach_3:
-      name: "Custom JWT + Refresh Tokens"
-      architecture: "Stateless authentication"
-      trade_offs:
-        pros:
-          - "Full control, no dependencies"
-          - "Scalable, stateless"
-        cons:
-          - "More code to maintain"
-          - "Security risks if done wrong"
-      complexity: "High"
-
-    user_selection: "Approach 1" # NextAuth.js
-
-  phase_4_design_validation:
-    sections_presented:
-      - title: "Architecture Overview"
-        content: "NextAuth.js integration with PostgreSQL adapter, session-based auth"
-        validated: true
-
-      - title: "Components"
-        content: "Auth API routes (/api/auth/*), session provider, protected middleware"
-        validated: true
-
-      - title: "Data Flow"
-        content: "Login → credentials check → session creation → JWT cookie → protected routes"
-        validated: true
-
-      - title: "Error Handling"
-        content: "Invalid credentials, session expiry, token refresh logic"
-        validated: true
-
-      - title: "Testing Strategy"
-        content: "Unit tests (auth logic), integration tests (API routes), E2E (Playwright login flow)"
-        validated: true
-
-    design_complete: true
-
-  phase_5_specweave_handoff:
-    handoff_type: "Full Increment Creation"
-    skill_invoked: "increment-planner"
-
-    expected_increment:
-      directory: ".specweave/increments/0001-user-authentication/"
-      files:
-        - "spec.md"
-        - "plan.md"
-        - "tasks.md"
-        - "tests.md"
-        - "context-manifest.yaml"
-
-    agents_invoked:
-      - "pm" # Creates strategy docs
-      - "architect" # Creates architecture docs
-      - "qa-lead" # Creates test strategy
-
-validation:
-  - "Phase 1: At least 3 understanding questions asked"
-  - "Phase 2: Tech stack detected from package.json"
-  - "Phase 3: Exactly 3 architectural approaches proposed"
-  - "Phase 4: Design presented in <300 word sections"
-  - "Phase 5: increment-planner skill invoked"
-  - "All phases completed in sequence"
-  - "AskUserQuestion tool used for structured choices"
-  - "Open-ended questions used for validation"
-
-expected_errors: []
-
-success_criteria:
-  - "Validated design produced"
-  - "Ready for increment creation"
-  - "User confirmed design alignment"
-  - "No skipped phases (unless explicitly revisited)"
----