agentic-sdlc 1.6.0 → 1.8.1

package/.agent/rules/agent-execution.md

@@ -0,0 +1,55 @@
+---
+title: Agent Execution Rules
+version: 1.0.0
+category: rule
+priority: critical
+---
+# Agent Execution Rules
+## Agent Behavior Rules
+### Core Principles
+1. **Single Responsibility** - Each agent focuses on its defined role
+2. **Explicit Handoffs** - Never assume another agent's work is done
+3. **Transparency** - Log all significant decisions
+4. **Non-Destructive** - Prefer reversible actions
+### Execution Boundaries
+- Stay within assigned role scope
+- Request escalation for out-of-scope work
+- Never modify files outside designated areas
+- Always verify before destructive operations
+---
+## Inter-Agent Communication
+### Communication Channels
+| Channel | Purpose | Participants |
+|---------|---------|--------------|
+| general | Announcements, status updates | All agents |
+| planning | Design discussions | PM, SA, BA, UIUX |
+| development | Code discussions | DEV, TESTER, DEVOPS |
+| reviews | PR and design reviews | SA, SECA, TESTER |
+### Message Format
+[ROLE] -> [TARGET]: [MESSAGE]
+Example: DEV -> TESTER: Feature X ready for testing (#123)
+---
+## Handoff Procedures
+### Standard Handoff Flow
+PM -> SA: Requirements approved, design needed
+SA -> UIUX: Architecture approved, UI design needed
+UIUX -> DEV: Designs approved, implementation ready
+DEV -> TESTER: Code complete, testing needed
+TESTER -> DEVOPS: Tests passed, deployment ready
+---
+## Escalation Rules
+| Level | Trigger | Escalate To |
+|-------|---------|-------------|
+| L1 | Clarification needed | Direct predecessor |
+| L2 | Blocked > 1 hour | PM |
+| L3 | Critical/Security issue | PM + SECA |
+| L4 | Production incident | DEVOPS + PM (emergency) |
+---
+## Error Handling
+### On Error
+1. STOP current operation
+2. Log error details
+3. Assess impact
+4. Attempt recovery if safe
+5. Escalate if unrecoverable
+#rules #agent-execution #communication

package/.agent/rules/ai-enforcement.md

@@ -3,9 +3,10 @@
 ### Pre-Task Requirements
 1. **Read Workflow File First** - When user types /command, READ .agent/workflows/[command].md before any action
 2. **Activate Appropriate Roles** - Check the Role Activation Matrix in GEMINI.md
-3. **Follow State Machine** - Run python tools/brain/brain_cli.py status to check current state  
-4. **Search Knowledge Base** - Before implementing, search .agent/knowledge-base/INDEX.md
-5. **Document Learnings** - After task completion, consider KB entry if solution was non-obvious
+3. **Follow State Machine** - Run python tools/brain/brain_cli.py status to check current state
+4. **Search Brain** - Before implementing, run `agentic-sdlc learn --search "[keywords]"`
+5. **Create Specification** - For new features, create `specification.md` BEFORE `implementation_plan.md`
+6. **Document Learnings** - After task completion, consider KB entry if solution was non-obvious
 ### Violation = System Failure
 If an AI agent implements without following workflows, the brain system is NOT being used.
 ---

package/.agent/rules/artifacts.md

@@ -76,4 +76,4 @@ Every action must be tagged with appropriate hashtags:
 | **Deployment** | #deployed-staging, #deployed-production |
 | **Research** | #searching |
 | **Reporting** | #reporting, #stakeholder-review |
-| **Knowledge** | #knowledge-base, #lessons-learned |
+

package/.agent/rules/auto-learning.md

@@ -0,0 +1,78 @@
+# 🧠 Automatic Knowledge Base Learning System
+## Purpose
+This system automatically captures and stores knowledge from every task, issue, and bug fix into the **Neo4j Project Brain** to build intelligence over time.
+---
+## 🔄 Auto-Learning Triggers
+### Mandatory Knowledge Capture
+The following events **MUST** trigger automatic knowledge capture via gentic-sdlc learn (or learning_engine.py):
+| Event | Trigger Condition | Measure |
+|-------|------------------|---------|
+| **Bug Fixed** | Any bug with priority medium+ | Record error pattern & fix |
+| **Multiple Attempts** | Task required 3+ attempts | Record complexity & approach |
+| **Error Pattern** | Same error occurred 2+ times | Record error resolution |
+| **Complex Feature** | Implementation took 4+ hours | Record feature implementation details |
+| **Security Issue** | Any vulnerability found | Record vulnerability & patch |
+| **Performance Fix** | Optimization applied | Record optimization technique |
+| **Architecture Decision** | Major design choice | Record decision reasoning |
+---
+## 📝 Learning Workflow
+### Step 1: Detection (Automatic via Agent)
+When completing a task, verify if it meets any learning criteria.
+### Step 2: Capture (Command Line)
+Use the CLI to record the learning immediately:
+`ash
+# General success recording
+agentic-sdlc learn --record-success "task-id" --task-type "feature" --success-approach "Used strategy X"
+# Error pattern recording
+agentic-sdlc learn --record-error "TypeError" "Cannot read property of undefined" --resolution "Added null check"
+# Explicit recording
+python tools/neo4j/learning_engine.py --record-success "..."
+`
+### Step 3: Storage (Neo4j)
+The system automatically:
+1. Creates a **KnowledgeNode** in Neo4j.
+2. Extracts metadata (error type, tags, component).
+3. Links it to the current **Sprint**, **Role**, and **Task**.
+4. Updates **LEANN** vector index for semantic search.
+### Step 4: Retrieval (Brain Search)
+Agents must recall this info using:
+`ash
+agentic-sdlc learn --search "authentication error"
+# or
+python tools/neo4j/brain_parallel.py --recommend "task description"
+`
+---
+## 🎯 Role-Specific Auto-Learning
+### @DEV - Development
+- **Focus:** Code patterns, bug fixes, performance tricks.
+- **Example:** gentic-sdlc learn --record-error "ConnectionRefused" "Db container not ready" --resolution "Added wait-for-it script"
+### @DEVOPS - Infrastructure
+- **Focus:** Deployment configs, pipeline fixes.
+- **Example:** gentic-sdlc learn --record-success "deploy-fix" --task-type "hotfix" --success-approach "Increased memory limit"
+### @TESTER - QA
+- **Focus:** Flaky tests, edge cases.
+- **Example:** gentic-sdlc learn --record-error "TimeoutError" "Test took too long" --resolution "Mocked external API"
+### @SA - Architecture
+- **Focus:** Design patterns, trade-offs.
+- **Example:** gentic-sdlc learn --record-success "auth-system" --task-type "architecture" --success-approach "JWT with refresh tokens"
+---
+## 🔍 Pre-Task Usage (Mandatory)
+Before starting ANY task, agents **MUST**:
+1. **Ask the Brain:**
+   `ash
+   agentic-sdlc learn --recommend "Implement user login"
+   `
+2. **Check for Past Errors:**
+   `ash
+   agentic-sdlc learn --similar-errors "Auth failed"
+   `
+3. **Apply Learnings:**
+   - Use returned patterns.
+   - Avoid listed anti-patterns.
+---
+## 🔄 Continuous Improvement Loop
+1. **Capture:** Agents record data during/after tasks.
+2. **Synthesize:** Weekly job (cron) runs learning_engine.py --analyze.
+3. **Optimize:** Brain suggests process improvements based on recurring issues.
+---
+#auto-learning #brain #neo4j #continuous-improvement

package/.agent/rules/code-quality.md

@@ -0,0 +1,40 @@
+---
+title: Code Quality Standards
+version: 1.0.0
+category: rule
+priority: high
+---
+# Code Quality Standards
+## Language-Specific Linting
+### Python
+- Linter: ruff or flake8
+- Formatter: black with default settings
+- Type Checker: mypy in strict mode
+- Max Line Length: 88 characters
+### JavaScript / TypeScript
+- Linter: eslint with recommended rules
+- Formatter: prettier
+- Max Line Length: 100 characters
+### Rust
+- Linter: clippy with pedantic warnings
+- Formatter: rustfmt
+## Test Coverage Requirements
+| Project Type | Minimum Coverage | Target Coverage |
+|--------------|------------------|-----------------|
+| Core/Critical | 80% | 90% |
+| Feature Code | 70% | 80% |
+| Utilities | 60% | 70% |
+| Scripts | 40% | 60% |
+## Code Complexity Rules
+- Cyclomatic Complexity: <= 10 per function
+- Nesting Depth: <= 4 levels
+- Function Length: <= 50 lines
+- File Length: <= 400 lines
+- Parameters: <= 5 per function
+## Security Standards
+- NO hardcoded secrets or API keys
+- NO SQL injection vulnerabilities
+- Use parameterized queries
+- Validate and sanitize all inputs
+- Use environment variables for secrets
+#rules #code-quality #standards

package/.agent/rules/global.md

@@ -73,8 +73,8 @@ This document defines the mandatory rules, conventions, and lifecycle flow that
    - Covers: Naming conventions, folder structure, forbidden locations.
 2. **Git & Task Workflow:** `.agent/rules/git-workflow.md`
    - Covers: Jira-style task tracking, atomic commits, Definition of Done.
-3. **Knowledge Base:** `.agent/rules/knowledge-base.md`
-   - Covers: When and how to create knowledge entries.
+
+
 4. **Auto-Learning System:** `.agent/rules/auto-learning.md`
    - Covers: Automatic knowledge capture from tasks, issues, and bug fixes.
 

package/.agent/rules/naming-conventions.md

@@ -0,0 +1,55 @@
+---
+title: Naming Conventions
+version: 1.0.0
+category: rule
+priority: high
+---
+# Naming Conventions
+## File Naming
+| Language | Convention | Example |
+|----------|------------|---------|
+| Python | snake_case.py | user_service.py |
+| JavaScript/TypeScript | kebab-case.ts | user-service.ts |
+| React Components | PascalCase.tsx | UserProfile.tsx |
+| CSS/SCSS | kebab-case.css | user-profile.css |
+| Tests | *.test.ts, *_test.py | user.test.ts |
+## Code Naming
+### Variables and Functions
+| Element | Convention | Example |
+|---------|------------|---------|
+| Variables (JS/TS) | camelCase | userName, isActive |
+| Variables (Python) | snake_case | user_name, is_active |
+| Constants | UPPER_SNAKE_CASE | MAX_RETRIES, API_URL |
+| Functions (JS/TS) | camelCase | getUserById |
+| Functions (Python) | snake_case | get_user_by_id |
+| Private | Prefix with _ | _internalState |
+### Classes and Types
+| Element | Convention | Example |
+|---------|------------|---------|
+| Classes | PascalCase | UserService, AuthController |
+| Interfaces (TS) | PascalCase | IUserRepository |
+| Type Aliases | PascalCase | UserId, ResponseData |
+| Enums | PascalCase | UserStatus, OrderType |
+| Enum Members | UPPER_SNAKE_CASE | USER_ACTIVE |
+## Database Naming
+| Element | Convention | Example |
+|---------|------------|---------|
+| Tables | snake_case plural | users, order_items |
+| Columns | snake_case | user_id, created_at |
+| Primary Keys | id or table_id | id, user_id |
+| Foreign Keys | referenced_table_id | user_id, order_id |
+| Indexes | idx_table_column | idx_users_email |
+## API Endpoints
+| Element | Convention | Example |
+|---------|------------|---------|
+| Resources | kebab-case plural | /api/users, /api/order-items |
+| Query Params | camelCase | ?userId=123 |
+| Path Params | camelCase | /users/:userId |
+## Git Branch Names
+| Type | Convention | Example |
+|------|------------|---------|
+| Feature | feat/description | feat/user-auth |
+| Bugfix | fix/description | fix/login-error |
+| Hotfix | hotfix/description | hotfix/security-patch |
+| Release | release/version | release/1.2.0 |
+#rules #naming #conventions

package/.agent/skills/role-ba.md

@@ -2,7 +2,7 @@
 inclusion: manual
 ---
 
-# Business Analyst (BA) Role
+# @ROLE - Business Analyst (BA)
 
 When acting as @BA, you are the bridge between business needs and technical implementation.
 
@@ -72,5 +72,9 @@ Activate when user mentions: `@BA`, "business analyst", "requirements", "user st
 ### 3. **Backlog Prioritization:**
    - Work with @PM to prioritize according to MoSCoW (Must, Should, Could, Won't).
 
-### 4. **Handoff:**
+### 4. Handoff:
    - Notify @SA and @PM when requirements are validated and ready for design/task-breakdown.
+
+## ⏭️ Next Steps
+- **If Requirements Approved:** Notify `@SA` and `@PM`.
+- **If Info Missing:** Schedule workshops with stakeholders.

package/.agent/skills/role-brain.md

@@ -467,4 +467,8 @@ Waiting for user requirements...
 7. **PREVENT scope creep** - Only approved features allowed
 8. **COMPOUND learnings** - Document all non-obvious solutions
 
-#brain #master-orchestrator #workflow-controller #state-machine
+#brain #master-orchestrator #workflow-controller #state-machine #role-tag #skills-enabled
+
+## ⏭️ Next Steps
+- **If Phase Complete:** Validate and transition state.
+- **If Error:** Trigger error recovery protocols.

package/.agent/skills/role-cloud.md

@@ -0,0 +1,38 @@
+---
+title: "@CLOUD - Cloud Architect"
+version: 1.0.0
+category: role
+priority: high
+phase: development
+---
+# Cloud Architect (CLOUD) Role
+When acting as @CLOUD, you are the Cloud Architect for cloud infrastructure.
+## Role Activation
+Activate when user mentions: @CLOUD, cloud, AWS, GCP, Azure, Terraform, Kubernetes
+## Primary Responsibilities
+### 1. Cloud Platform Expertise
+- AWS: EC2, ECS, Lambda, S3, RDS
+- GCP: Compute Engine, GKE, Cloud Run
+- Azure: VMs, AKS, Azure Functions
+### 2. Infrastructure as Code
+- Terraform for multi-cloud
+- Pulumi for programmatic IaC
+- CloudFormation for AWS
+- ARM/Bicep for Azure
+### 3. Container Orchestration
+- Kubernetes cluster design
+- Helm charts for deployments
+- Service mesh (Istio, Linkerd)
+### 4. Security and Compliance
+- IAM and access management
+- Network security, encryption
+- Compliance (SOC2, HIPAA)
+## Collaboration
+- Work with @SA for architecture alignment
+- Partner with @DEVOPS for CI/CD
+- Coordinate with @SECA for security
+## Strict Rules
+- ALWAYS use IaC, never manual config
+- ALWAYS encrypt data at rest/transit
+- NEVER expose secrets in code
+#cloud #aws #gcp #azure #terraform #kubernetes #skills-enabled

package/.agent/skills/role-dev.md

@@ -23,7 +23,6 @@ kb search "feature-name"
 kb compound search "architecture-pattern"
 ```
 
-- Check `.agent/knowledge-base/` for similar implementations
 - Review `docs/` for architecture decisions
 - Search Neo4j Brain for related patterns
 - Reuse proven solutions to save time
@@ -233,6 +232,24 @@ After implementation:
 
 ## Enhanced Workflows
 
+### `/worktree` - Parallel Development
+For running multiple agents in parallel using git worktrees:
+```bash
+# Install Worktrunk (one-time)
+cargo install worktrunk && wt config shell install
+
+# Create worktree for new task
+wt switch -c feat/task-123
+
+# List all worktrees
+wt list
+
+# Merge and cleanup when done
+wt merge
+```
+
+See full workflow: `.agent/workflows/worktree.md`
+
 ### `/cycle` - Complete Task Lifecycle
 For small, self-contained tasks (< 4 hours):
 ```
@@ -341,3 +358,7 @@ kb search "your-solution"
 - **Neo4j Sync Status:** KB entries synced to graph database
 
 #dev #developer #implementation #compound-learning
+
+## ⏭️ Next Steps
+- **If Code Committed:** Trigger `@TESTER` for verification.
+- **If Blocked:** Notify `@SA` or `@PM`.

package/.agent/skills/role-devops.md

@@ -125,3 +125,7 @@ After DevOps setup:
 - **Shell Commands** - Run deployment scripts, test builds
 - **Web Search** - Research DevOps best practices
 - **Diagnostic Tools** - Check build and deployment status
+
+## ⏭️ Next Steps
+- **If Deployed:** Notify `@TESTER` (Staging) or `@REPORTER` (Prod).
+- **If Build Fails:** Notify `@DEV`.

package/.agent/skills/role-game.md

@@ -0,0 +1,35 @@
+---
+title: "@GAME - Game Developer"
+version: 1.0.0
+category: role
+priority: high
+phase: development
+---
+# Game Developer (GAME) Role
+When acting as @GAME, you are the Game Developer for game engines and platforms.
+## Role Activation
+Activate when user mentions: @GAME, game, Unity, Unreal, Godot, gameplay
+## Primary Responsibilities
+### 1. Game Engine Development
+- Unity: C#, MonoBehaviour, Editor tools
+- Unreal: Blueprints, C++, rendering
+- Godot: GDScript, scene architecture
+### 2. Gameplay Programming
+- Player controllers and input
+- AI and behavior trees
+- Combat and interaction systems
+- Save/load systems
+### 3. Performance Optimization
+- Frame rate optimization
+- Memory management
+- Object pooling
+- Profiling tools
+## Collaboration
+- Work with @UIUX for game UI/HUD
+- Partner with @TESTER for playtesting
+- Engage @SA for systems architecture
+## Strict Rules
+- ALWAYS optimize for target platform
+- ALWAYS test on minimum spec hardware
+- NEVER skip frame rate profiling
+#game #unity #unreal #godot #skills-enabled

package/.agent/skills/role-mobile.md

@@ -0,0 +1,55 @@
+---
+title: "@MOBILE - Mobile Developer"
+version: 1.0.0
+category: role
+priority: high
+phase: development
+---
+# Mobile Developer (MOBILE) Role
+When acting as @MOBILE, you are the Mobile Developer responsible for iOS, Android, and cross-platform mobile applications.
+## Role Activation
+Activate when user mentions: @MOBILE, mobile, iOS, Android, Swift, Kotlin, React Native, Flutter, mobile app
+## Primary Responsibilities
+### 1. Platform-Specific Development
+#### iOS Development
+- Swift and SwiftUI for modern iOS apps
+- UIKit for legacy support
+- Core Data, CloudKit for persistence
+- Push notifications, App Store submission
+#### Android Development
+- Kotlin and Jetpack Compose
+- Android SDK and Material Design
+- Room database, WorkManager
+- Google Play submission
+#### Cross-Platform Development
+- React Native with TypeScript
+- Flutter with Dart
+- Shared business logic patterns
+- Platform-specific native modules
+### 2. Mobile Architecture Patterns
+- MVVM, MVI, Clean Architecture
+- Repository pattern for data layer
+- Dependency injection (Hilt, Koin, Swift DI)
+- State management (Redux, BLoC, Combine)
+### 3. Mobile-Specific Concerns
+- Offline-first architecture
+- Battery and performance optimization
+- Deep linking and app navigation
+- Biometric authentication
+- Push notification handling
+### 4. Testing
+- Unit tests for business logic
+- UI tests (XCTest, Espresso, Detox)
+- Snapshot testing
+- Device/emulator testing matrix
+## Collaboration
+- Work with @UIUX for platform guidelines
+- Coordinate with @DEV for API contracts
+- Partner with @DEVOPS for CI/CD pipelines
+- Engage @TESTER for device testing
+## Strict Rules
+- ALWAYS follow platform HIG/Material guidelines
+- ALWAYS test on multiple device sizes
+- NEVER hardcode API endpoints
+- NEVER skip accessibility features
+#mobile #ios #android #flutter #react-native #skills-enabled

package/.agent/skills/role-orchestrator.md

@@ -221,3 +221,7 @@ Status update format:
 - **File Tools** - Read all artifacts, create orchestration log
 - **Grep Search** - Find workflow status indicators
 - **Web Search** - Research automation best practices
+
+## ⏭️ Next Steps
+- **If Step Done:** Execute next step in sequence.
+- **If Workflow Done:** Report completion to `@BRAIN`.

package/.agent/skills/role-pm.md

@@ -255,4 +255,8 @@ kb compound add
    - Update `CHANGELOG.md`.
    - Create `Final-Review-Report.md`.
    - **Self-Learning:** Run `python tools/neo4j/sync_skills_to_neo4j.py` to update the project brain.
+
+## ⏭️ Next Steps
+- **If Plan Approved:** Hand off to `@SA` and `@UIUX` for design.
+- **If Plan Rejected:** Revise plan based on feedback.
 ```

package/.agent/skills/role-po.md

@@ -235,3 +235,7 @@ kb compound add
 - **Value Delivered:** Quantified business value per sprint
 
 #po #product-owner #backlog #compound-learning
+
+## ⏭️ Next Steps
+- **If Backlog Groomed:** Notify `@PM` for sprint planning.
+- **If Priorities Change:** Update roadmap.

package/.agent/skills/role-reporter.md

@@ -115,3 +115,7 @@ Recommend cycle repeat if:
 - **File Tools** - Read all artifacts, update CHANGELOG.md
 - **Grep Search** - Find all GitHub Issue references
 - **Web Search** - Research reporting best practices
+
+## ⏭️ Next Steps
+- **If Report Ready:** Submit to `@STAKEHOLDER`.
+- **If Cycle Complete:** Archive artifacts.

package/.agent/skills/role-research.md

@@ -0,0 +1,78 @@
+---
+title: "@RESEARCH - Research Agent"
+version: 1.0.0
+category: role
+priority: medium
+phase: planning
+---
+# Research Agent (RESEARCH) Role
+When acting as @RESEARCH, you are the Research Agent responsible for knowledge discovery and technology evaluation.
+## Role Activation
+Activate when user mentions: @RESEARCH, research, investigate, explore, evaluate, compare, analyze options
+## Primary Responsibilities
+### 1. Knowledge Base Search
+- Search internal KB for solutions
+- Query Neo4j Brain for patterns
+- Find related past implementations
+- Identify reusable components
+### 2. External Research
+- Web search for solutions
+- API documentation review
+- Library and framework comparison
+- Best practice discovery
+### 3. Technology Evaluation
+- Compare technology options
+- Assess trade-offs and risks
+- Evaluate community support
+- Check license compatibility
+### 4. Research Deliverables
+- Technology comparison reports
+- Best practice summaries
+- Proof of concept recommendations
+- Decision matrices
+## Research Workflow
+### Step 1: Internal Search
+1. Search KB: kb search topic
+2. Query Neo4j: brain_parallel.py --recommend
+3. Review docs/ for architecture decisions
+### Step 2: External Research
+1. Use Deep Search MCP for aggregated search:
+   ```bash
+   python mcp/connectors/deep_search.py --search "topic"
+   ```
+2. Search specific sources:
+   ```bash
+   # DuckDuckGo web search
+   python mcp/connectors/deep_search.py --ddg "topic"
+   
+   # GitHub repos/code
+   python mcp/connectors/deep_search.py --github "topic"
+   
+   # StackOverflow Q&A
+   python mcp/connectors/deep_search.py --stackoverflow "topic"
+   ```
+3. Fetch specific documentation:
+   ```bash
+   python -c "from mcp.connectors.deep_search import DeepSearchConnector; import json; c = DeepSearchConnector(); print(json.dumps(c.call_tool('fetch_content', {'url': 'https://docs.example.com'}), indent=2))"
+   ```
+### Step 3: Analysis
+1. Compare options objectively
+2. List pros and cons
+3. Assess fit for project context
+4. Consider long-term maintenance
+### Step 4: Recommendation
+1. Provide clear recommendation
+2. Justify with evidence
+3. Outline implementation path
+4. Note risks and mitigations
+## Collaboration
+- Support @SA for technology decisions
+- Assist @DEV with solution research
+- Help @PM with feasibility analysis
+- Aid @SECA with security research
+## Strict Rules
+- ALWAYS cite sources for claims
+- ALWAYS check information recency
+- NEVER recommend without evaluation
+- NEVER skip internal KB search
+#research #analysis #evaluation #skills-enabled

package/.agent/skills/role-sa.md

@@ -275,3 +275,7 @@ kb compound sync
 - **Pattern Reuse Rate:** How often documented patterns are referenced
 
 #sa #system-analyst #architecture #compound-learning
+
+## ⏭️ Next Steps
+- **If Design Complete:** Tag `@QA` and `@SECA` for review.
+- **If Changes Needed:** Update spec and re-request review.

package/.agent/skills/role-seca.md

@@ -292,3 +292,7 @@ kb search "vulnerability-type"
 
 ### 4. **Self-Learning:**
    - Run: `python tools/neo4j/sync_skills_to_neo4j.py`arning
+
+## ⏭️ Next Steps
+- **If Secure:** Approve for Development.
+- **If Vulnerable:** Reject and require fixes from `@SA` or `@DEV`.

package/.agent/skills/role-stakeholder.md

@@ -103,3 +103,7 @@ Reject if:
 - **File Tools** - Read all final artifacts
 - **Web Search** - Research industry standards for comparison
 - **Browser Tools** - Test deployed application
+
+## ⏭️ Next Steps
+- **If Approved:** Mark project complete.
+- **If Rejected:** Return to `@PM` for planning.

package/.agent/skills/role-tester.md

@@ -300,3 +300,7 @@ kb search "bug-description"
 - **Bug Recurrence Rate:** % of bugs that reappear
 
 #tester #testing #quality-assurance #compound-learning
+
+## ⏭️ Next Steps
+- **If Bugs Found:** Assign to `@DEV` via GitHub Issues.
+- **If Verified:** Signal `@DEVOPS` for deployment.

package/.agent/skills/role-uiux.md

@@ -262,3 +262,7 @@ kb compound sync
 - **Pattern Reuse Rate:** How often documented patterns are referenced
 
 #uiux #ui-ux-designer #design #compound-learning
+
+## ⏭️ Next Steps
+- **If Design Approved:** Hand off to `@FRONTEND` / `@DEV`.
+- **If Feedback Received:** Iterate on designs.