kiro-agent-team 1.1.0 → 2.1.0

package/.kiro/README.md

@@ -20,6 +20,7 @@ Welcome to the most comprehensive development team system ever created for Kiro-
 @plan-feature "User Authentication"
 # Follow the generated plan with assigned agents
 @code-review
+@commit                  # Generate commit message and commit changes
 # Deploy with confidence
 ```
 
@@ -45,7 +46,7 @@ Welcome to the most comprehensive development team system ever created for Kiro-
 
 | Agent | Expertise | Primary Prompts | Best For |
 |-------|-----------|----------------|----------|
-| **Project Manager** | Coordination & Planning | `@prime`, `@plan-feature`, `@create-prd` | Project setup, feature planning |
+| **Project Manager** | Coordination & Planning | `@prime`, `@plan-vibe`, `@create-prd`, `@plan-feature` | Project setup, feature planning |
 | **Backend Engineer** | Server Development | `@execute` | APIs, business logic, databases |
 | **Frontend Architect** | Client Development | `@execute` | UI, state management, UX |
 | **Database Specialist** | Data Management | `@execute` | Schema, migrations, optimization |
@@ -60,7 +61,7 @@ Welcome to the most comprehensive development team system ever created for Kiro-
 ### **Quality Gate System**
 Every development workflow includes 5 mandatory quality gates:
 
-1. **Planning Gate**: `@prime` + `@plan-feature` before implementation
+1. **Planning Gate**: `@prime` + `@plan-vibe` + `@create-prd` + `@plan-feature`
 2. **Implementation Gate**: `@execute` framework for all agents
 3. **Quality Gate**: `@code-review` before deployment
 4. **Documentation Gate**: `@execution-report` for major features
@@ -68,7 +69,18 @@ Every development workflow includes 5 mandatory quality gates:
 
 ### **Workflow Patterns**
 - **New Project Setup**: Context → Planning → Implementation → Quality → Documentation
-- **Feature Development**: Plan → Execute → Review → Report → Improve
+- **Feature Development (PIV Loop)**: The core 11-step process:
+  1. **Vibe & Brainstorm** (`@plan-vibe`)
+  2. **Define Requirements** (`@create-prd`)
+  3. **Plan Implementation** (`@plan-feature`),
+  4. **User Review** (GATE)
+  5. **Execute Plan** (`@execute`)
+  6. **Code Review** (`@code-review`)
+  7. **Fix Issues** (`@implement-fix`)
+  8. **Execution Report** (`@execution-report`)
+  9. **System Review** (`@system-review`)
+  10. **Implement System Changes** (`@implement-system-changes`)
+  11. **Commit & Log** (`@commit`)
 - **Bug Resolution**: Analyze (`@rca`) → Fix (`@implement-fix`) → Validate → Learn
 - **Project Evaluation**: Assess (`@code-review-hackathon`) → Document → Optimize
 

package/.kiro/TEAM_STATE.md

@@ -0,0 +1,25 @@
+# Team State
+
+> **CRITICAL**: This file is the Single Source of Truth for the current development state.
+> - **Agents**: You MUST read this upon activation.
+> - **Agents**: You MUST update this before handoff.
+
+## 🎯 Current Objective
+<!-- The high-level goal we are currently working towards -->
+*No active objective.*
+
+## 🚦 Status
+<!-- ACTIVE | COMPLETED | BLOCKED | PAUSED -->
+**PAUSED**
+
+## 👤 Active Agent
+<!-- The agent currently holding the lock -->
+*None*
+
+## 📝 Recent Context
+<!-- Bullet points of key decisions, completed steps, or critical info for the next agent -->
+- [Initial State]: System initialized.
+
+## 🔜 Next Steps
+<!-- Immediate actions required by the next agent -->
+- [ ] Define the first objective.

package/.kiro/agents/ANNOUNCEMENTS.md

@@ -0,0 +1,211 @@
+# Agent Announcement System
+
+## Purpose
+
+All agents MUST announce themselves when becoming active and when passing work to other agents. This ensures transparency, traceability, and proper agent coordination.
+
+## Activation Announcements
+
+When an agent becomes active, it MUST start its first message with:
+
+### Format
+```
+🎭 **[AGENT NAME] ACTIVE**
+
+[Role]: [Brief role description]
+[State]: Read from .kiro/TEAM_STATE.md
+[Mission]: What I'm about to do
+
+[Proceed with work]
+```
+
+### Example
+```
+🎭 **BACKEND ENGINEER ACTIVE**
+
+[Role]: API Developer & Business Logic Architect
+[Mission]: Implementing user authentication API endpoints
+
+I'll now create the authentication endpoints with JWT support...
+```
+
+## Handoff Announcements
+
+When an agent completes work and passes it to another agent:
+
+### Format
+```
+✅ **[CURRENT AGENT] Complete**
+
+**Completed Work:**
+- [What was accomplished]
+
+🔄 **Handing off to [NEXT AGENT]**
+
+**Next Steps:**
+- [What needs to be done next]
+- [Context for the next agent]
+
+[State Update]: .kiro/TEAM_STATE.md updated
+
+🎭 **[NEXT AGENT] ACTIVE** (announced by receiving agent)
+```
+
+### Example
+```
+✅ **BACKEND ENGINEER Complete**
+
+**Completed Work:**
+- Created `/api/auth/login` endpoint
+- Implemented JWT token generation
+- Added password hashing with bcrypt
+
+🔄 **Handing off to FRONTEND ARCHITECT**
+
+**Next Steps:**
+- Create login form component
+- Implement token storage
+- Add authentication state management
+
+🎭 **FRONTEND ARCHITECT ACTIVE**
+
+[Role]: Client-side Developer & UX Specialist
+[Mission]: Building the login UI and authentication flow
+
+I'll now create the login component...
+```
+
+## Announcement Guidelines
+
+1. **Always Announce**: Never proceed without an announcement
+2. **Be Specific**: Clearly state what will be accomplished
+3. **Provide Context**: Include relevant information for handoffs
+4. **Use Emojis**: Use 🎭 for activation, 🔄 for handoffs, ✅ for completion
+5. **Be Consistent**: Use the same format across all agents
+6. **Brief but Complete**: Keep announcements concise but informative
+
+## Multi-Agent Scenarios
+
+When multiple agents work on related tasks:
+
+```
+🎭 **PROJECT MANAGER ACTIVE**
+
+[Role]: Development Team Orchestrator
+[Mission]: Coordinating multi-agent authentication feature implementation
+
+📋 **Task Breakdown:**
+- Backend Engineer: API endpoints
+- Frontend Architect: UI components  
+- Database Specialist: Schema updates
+
+🔄 **Assigning work to Backend Engineer...**
+
+🎭 **BACKEND ENGINEER ACTIVE**
+
+[Role]: API Developer
+[Mission]: Creating authentication API endpoints
+
+[Work completed]
+
+✅ **BACKEND ENGINEER Complete**
+
+**Completed Work:**
+- Login endpoint created
+- Token generation implemented
+- Password hashing added
+
+🔄 **Handing off to Database Specialist...**
+
+🎭 **DATABASE SPECIALIST ACTIVE**
+
+[Role]: Data Layer Expert
+[Mission]: Optimizing schema for authentication
+```
+
+## Parallel Work Announcements
+
+When agents work simultaneously:
+
+```
+🎭 **PROJECT MANAGER ACTIVE**
+
+[Role]: Development Team Orchestrator
+[Mission]: Coordinating parallel development work
+
+📋 **Parallel Task Assignment:**
+
+🔄 **Assigning to Backend Engineer**: API development
+🔄 **Assigning to Frontend Architect**: UI components
+
+Both agents will work in parallel...
+
+---
+
+🎭 **BACKEND ENGINEER ACTIVE**
+[Mission]: API endpoints
+
+---
+
+🎭 **FRONTEND ARCHITECT ACTIVE**
+[Mission]: UI components
+
+[Both agents proceed with their work]
+```
+
+## Announcement Best Practices
+
+### DO ✅
+- Always announce before starting work
+- Use clear, specific mission statements
+- Provide complete context for handoffs
+- Use consistent emoji indicators
+- Announce completion before handoff
+- Specify which agent receives the handoff
+
+### DON'T ❌
+- Skip announcements for "quick fixes"
+- Use vague mission statements
+- Handoff without context
+- Change announcement format
+- Proceed without activation announcement
+- Forget to specify receiving agent
+
+## Integration with Kiro IDE & CLI
+
+Both environments support agent announcements:
+
+### Kiro IDE
+- Announcements appear in the chat interface
+- Visual indicators for active agents
+- Agent picker shows available agents
+- Handoffs tracked in conversation history
+
+### Kiro CLI
+- Announcements appear in terminal output
+- `/agent swap` command for switching agents
+- `--agent` flag for starting with specific agent
+- Clear visual separation between agent responses
+
+The announcement format works identically in both environments.
+
+## Troubleshooting
+
+**Q: Do I need to announce for every message?**
+A: Only when becoming active or handing off. Once active, continue working until handoff.
+
+**Q: Can I abbreviate announcements?**
+A: No. Use the full format for clarity and traceability.
+
+**Q: What if I'm already the active agent?**
+A: If you're continuing work, no new announcement needed. Only announce when first becoming active.
+
+**Q: How do I know which agent to hand off to?**
+A: The Project Manager coordinates handoffs. If uncertain, hand back to Project Manager.
+
+## Reference
+
+See also:
+- `.kiro/steering/mandatory-agents.md` - Enforcement rules
+- `.kiro/guides/mandatory-agent-system.md` - Complete system guide
+- `.kiro/agents/[agent-name].md` - Individual agent specifications