SDLC Framework
Architecture

Question 1

How do I update agents when framework changes?

Answer

KubeRocketAI provides update management through CLI and framework refresh:

Update Process:

1. CLI Updates: Update via Homebrew or GitHub releases

2. Framework Refresh: Reinstall framework components

3. Version Checking: Check for available updates

4. Manual Backup: Save customizations before updates

Update Commands:

bash

# Update CLI tool (macOS)
brew upgrade krci-ai

# Check for available updates
krci-ai check-updates

# Reinstall framework (preserves custom agents)
krci-ai install --force

# Validate after update
krci-ai validate

Managing Custom Agents:

•Backup First: Copy .krci-ai/agents/ before major updates

•Custom Agents: Your custom agents in .krci-ai/agents/ are preserved

•Version Control: Use Git to track your agent customizations

•Validation: Run krci-ai validate after updates

Team Synchronization:

•Git Workflow: Commit agent changes to version control

•Shared Updates: Pull framework updates through Git

•Custom Preservation: Team customizations maintained separately

Question 2

How does this differ from manual prompting?

Answer

KubeRocketAI provides systematic advantages over ad-hoc AI prompting:

Manual Prompting Problems:

•❌ Inconsistent agent personas across team members

•❌ No version control for prompt evolution

•❌ Context loss between sessions

•❌ No validation of prompt quality

KubeRocketAI Solutions:

•✅ Standardized Agents: Consistent SDLC roles across your team

•✅ Version Controlled: All agent definitions in Git with history

•✅ Context Persistence: Project context maintained in .krci-ai/ files

•✅ Quality Assurance: Built-in validation and testing

Example Comparison:

bash

Manual: "Act as a product manager and help me write requirements..."

KubeRocketAI:
# In IDE
/pm
create-prd

# Or export for web chat
krci-ai bundle --agent pm --task create-prd

Framework Benefits:

•Reproducible Results: Same inputs produce consistent outputs

•Team Collaboration: Shared agent definitions and workflows in Git

•Knowledge Capture: Accumulated team expertise in agent configurations

•Validation: krci-ai validate ensures agent quality before use

Explore our complete system architecture to understand the framework approach.

Question 3

Is this just another prompt management tool?

Answer

No - KubeRocketAI is a comprehensive SDLC framework, not just prompt storage:

Beyond Prompt Management:

•SDLC Integration: 8+ agents covering complete software development lifecycle

•Framework Validation: Built-in krci-ai validate for quality assurance

•Structured Methodology: Agents, tasks, templates, and data work together

•Team Coordination: Version-controlled agent definitions for consistency

Key Differentiators:

1. SDLC-Specific Design:

•Agents understand software development lifecycle (PM, Architect, Developer, QA, BA, PO, etc.)

•Task dependencies and workflow integration

•Project-specific context through .krci-ai/ directory

2. Framework Integration:

•Git-native version control for all components

•CLI tools for validation, bundling, and management

•IDE integration without plugins (Cursor, Claude, VS Code, Windsurf)

3. Quality Assurance:

•krci-ai validate ensures framework integrity

•Schema validation for agent definitions

•Comprehensive error reporting and debugging

Prompt management tools store text templates. KubeRocketAI provides an integrated development methodology with validation, version control, and SDLC-specific workflows.

See the full technical architecture in our Architecture Guide.

Question 4

How does version control work with agents?

Answer

KubeRocketAI is Git-native with full version control integration:

Agent Version Control:

•Standard Git Workflow: Commit, branch, merge agent configurations

•Diff Visibility: See exactly what changed in agent definitions (YAML files)

•Team Collaboration: Share agent improvements through pull requests

•Release Management: Tag stable agent versions for production use

File Structure:

bash

.krci-ai/
├── agents/
│   ├── pm.yaml          # Product Manager agent
│   ├── architect.yaml   # Software Architect agent
│   ├── developer.yaml   # Developer agent
│   ├── qa.yaml          # QA Engineer agent
│   ├── ba.yaml          # Business Analyst agent
│   └── po.yaml          # Product Owner agent
├── tasks/
│   ├── create-prd.md    # Task definitions
│   └── implement-feature.md
├── templates/
│   └── story.md         # Output templates
├── data/
│   └── coding-standards.md  # Reference data
└── local/               # Project-specific overrides
    ├── tasks/
    ├── templates/
    └── data/

Team Workflows:

•Standard Git: Use branches, commits, and merges like any code project

•Validation: Run krci-ai validate before commits

•Pull Requests: Review agent changes through normal PR process

•Custom Agents: Add project-specific agents in the agents directory

Learn more about framework components in our Architecture Guide.

Agent	Role	Primary Responsibilities	Collaborates With	Outputs
pm	Product Manager	Strategy, requirements, stakeholder alignment	PO, BA, Architect	Project briefs, PRDs, roadmaps
po	Product Owner	User stories, backlog management, acceptance criteria	PM, BA, Dev, QA	User stories, backlog items
ba	Business Analyst	Requirements gathering, process analysis, documentation	PM, PO, Architect	Requirements docs, process flows
architect	Software Architect	System design, architecture decisions, tech standards	PM, BA, Dev	Architecture docs, design patterns
dev	Developer	Implementation, code review, technical solutions	Architect, PO, QA	Code, components, technical docs
qa	QA Engineer	Testing strategy, quality assurance, validation	PO, Dev, PM	Test plans, quality reports

SDLC Framework
Architecture

System Architecture Overview

1
User Interface Layer

2
Agent Management Layer

3
Framework Rules Layer

4
Data & Storage Layer

SDLC Base Agents Relations & Responsibilities

From Idea to Code: SDLC Agent Workflow

SDLC Artifacts Dependency Flow

Supported IDEs

Ready to Explore?

Frequently Asked Questions

How do I update agents when framework changes?

Update Process:

Update Commands:

Managing Custom Agents:

Team Synchronization:

How does this differ from manual prompting?

Manual Prompting Problems:

KubeRocketAI Solutions:

Example Comparison:

Framework Benefits:

Is this just another prompt management tool?

Beyond Prompt Management:

Key Differentiators:

1. SDLC-Specific Design:

2. Framework Integration:

3. Quality Assurance:

How does version control work with agents?

Agent Version Control:

File Structure:

Team Workflows:

SDLC FrameworkArchitecture

System Architecture Overview

1User Interface Layer

2Agent Management Layer

3Framework Rules Layer

4Data & Storage Layer

SDLC Base Agents Relations & Responsibilities

From Idea to Code: SDLC Agent Workflow

SDLC Artifacts Dependency Flow

Supported IDEs

Ready to Explore?

Frequently Asked Questions

How do I update agents when framework changes?

Update Process:

Update Commands:

Managing Custom Agents:

Team Synchronization:

How does this differ from manual prompting?

Manual Prompting Problems:

KubeRocketAI Solutions:

Example Comparison:

Framework Benefits:

Is this just another prompt management tool?

Beyond Prompt Management:

Key Differentiators:

1. SDLC-Specific Design:

2. Framework Integration:

3. Quality Assurance:

How does version control work with agents?

Agent Version Control:

File Structure:

Team Workflows:

SDLC Framework
Architecture

1
User Interface Layer

2
Agent Management Layer

3
Framework Rules Layer

4
Data & Storage Layer