Forem: Ritika Srivastava

Notion challenge

Ritika Srivastava — Fri, 06 Mar 2026 13:18:45 +0000

DevFlow : AI-Powered Development Workflow Automation

Ritika Srivastava — Sun, 15 Feb 2026 12:21:16 +0000

This is a submission for the [GitHub Copilot CLI Challenge(https://github.com/sridurgeshv/GitHub-Copilot-CLI-Challenge)

What I Built

DevFlow is an AI-powered CLI that transforms how developers work with the terminal. It understands natural language commands and automates the entire development workflow—from git commits to deployments. Built on Python, it combines Google Gemini 2.0 with a robust command execution engine to handle tasks that typically require memorizing dozens of CLI tools and commands.

The killer feature is the Hybrid AI Code Review system—the first CLI tool to combine GitHub Copilot CLI + Google Gemini for comprehensive code analysis. Instead of relying on a single AI that might miss security issues or architectural problems, DevFlow uses both to catch what single solutions miss: code-level bugs (Copilot), security vulnerabilities, and architectural flaws (Gemini).

This project means everything to me because it solves a real pain point I experience daily—the cognitive overhead of remembering complex commands. It empowers junior developers to work with the confidence of seniors, and it makes deploying production code as simple as asking for it in plain English.

Demo

https://youtu.be/U4m7NOV-Gxs

My Experience with GitHub Copilot CLI

GitHub Copilot CLI dramatically accelerated DevFlow's development and shaped its core functionality. Here's how it impacted my workflow:

🎯 How I Used It:

Intent Recognition & Parsing: Copilot CLI helped me understand patterns in how developers naturally describe tasks. I used its analysis capabilities to inform DevFlow's natural language processing engine, ensuring DevFlow commands feel intuitive and conversational.
Code Analysis & Suggestions: While building the code review feature, Copilot CLI's analysis capabilities inspired and informed the hybrid review system. I studied how it identifies code patterns and integrated similar logic into DevFlow's analysis pipeline.
Shell Command Generation: Copilot CLI's command suggestion feature helped me understand the best patterns for safely executing complex shell commands. This directly influenced how I built DevFlow's executor module with safety confirmations and dry-run modes.
Error Handling & Recovery: Using Copilot CLI's error messages and recovery suggestions helped me design better error handling in DevFlow, with context-aware hints and actionable next steps.

📊 Impact on Development Experience:

70% Faster Prototyping: Instead of writing boilerplate code from scratch, Copilot CLI suggestions accelerated the initial architecture design
Better Safety Practices: Learning from Copilot CLI's safety-first approach led me to implement confirmation dialogs and dry-run capabilities in DevFlow
Clearer Intent Patterns: Copilot CLI's command parsing helped me structure DevFlow's natural language understanding around patterns developers already use
Production-Ready Code: The quality of Copilot CLI's suggestions meant fewer debugging cycles and more confidence in the final product

DevFlow essentially uses GitHub Copilot CLI as both inspiration AND integration point—the hybrid code review feature specifically leverages Copilot CLI's code analysis strength combined with Gemini's security expertise.

Team members:
sridurgesh
Ritika Srivastava

Building Graveyard Shift: How Kiro Changed My Development Workflow

Ritika Srivastava — Thu, 04 Dec 2025 08:11:09 +0000

Building Graveyard Shift: How Kiro Changed My Development Workflow

When I started building Graveyard Shift - a Halloween-themed task management app - I had no idea it would fundamentally change how I approach development. This is the story of how Kiro transformed my workflow from chaotic coding to structured, confident building.

The Idea: Spooky Productivity

The concept was simple: what if your todo list lived in a graveyard? Tasks become tombstones, goals become cursed dolls, and your productivity is tracked by moon phases. It's quirky, it's fun, and it needed to be polished.

But here's the thing - I wanted to build it fast without sacrificing quality. That's where Kiro came in.

The Old Way vs. The Kiro Way

Before Kiro

My typical workflow looked like this:

Start coding immediately
Realize I forgot something important
Refactor everything
Break something else
Spend hours debugging
Repeat

With Kiro

The workflow became:

Spec it out - Define requirements, design, and tasks
Build iteratively - Implement feature by feature
Get instant feedback - Kiro catches issues immediately
Refactor fearlessly - Context awareness means no broken connections
Ship confidently - Everything works together

The Spec-Driven Approach

One of Kiro's killer features is the spec system. Instead of diving into code, I started by creating structured specifications:

Requirements (EARS Syntax)

WHEN a user completes a task
THEN the system SHALL update the task status to 'completed'
AND the system SHALL record the completion timestamp
AND the system SHALL update the weekly progress moon phase

This forced me to think through the entire feature before writing a single line of code. Game changer.

Design (Property-Based Testing)

Property: Task Status Consistency
GIVEN any task in the system
WHEN the task status is updated
THEN the status MUST be one of: pending, completed, or overdue
AND the status MUST be reflected consistently across all views

By defining properties upfront, I knew exactly what needed to be tested and what "correct" looked like.

Real Examples: Features That Came Together Fast

1. Pomodoro Timer Integration

I wanted users to track focus sessions alongside their tasks. With Kiro:

Defined the spec in 10 minutes
Implemented the backend routes with instant validation
Built the React component with real-time feedback
Integrated with the existing task system seamlessly

Total time: ~2 hours for a fully working feature with backend persistence.

2. 3D Tombstone Cemetery

The visual centerpiece of the app - interactive 3D tombstones that users can click to view task details.

The challenge? Making it accessible, performant, and gracefully degrading for older browsers.

Kiro helped me:

Implement keyboard navigation (arrow keys, Enter, Escape)
Add ARIA labels for screen readers
Detect 3D transform support and fallback gracefully
Handle focus management for modals

All while maintaining the spooky aesthetic.

3. Cursed Dolls (Goal Tracking)

This was the most complex feature - long-term goals with progress tracking, milestone management, and visual feedback.

What would have taken days took hours because:

Kiro understood the entire codebase context
Suggested consistent patterns from existing features
Caught edge cases I would have missed
Helped refactor without breaking anything

The Technical Stack

Frontend:

React 19 with React Router
Framer Motion for animations
Three.js for 3D effects
Tailwind CSS + custom spooky styling

Backend:

Node.js + Express
MongoDB with Mongoose
JWT authentication
Appwrite for user management

The Kiro Advantage:
Kiro kept everything consistent across the stack. When I updated the task schema in the backend, it helped me update all the frontend components that depended on it. No more hunting through files wondering what I broke.

Lessons Learned

1. Specs Save Time

Writing specs felt slow at first, but it saved hours of refactoring later. Knowing exactly what you're building before you build it is powerful.

2. Context Awareness is Everything

Not having to explain my codebase repeatedly was liberating. Kiro just knew what I was working with.

3. Iteration Speed Matters

The faster you can try ideas, the better your final product. Kiro's instant feedback loop meant I could experiment without fear.

4. Quality and Speed Aren't Opposites

I shipped faster and with better code quality. Property-based testing, accessibility compliance, and error handling all came naturally.

The Results

In less time than I'd normally spend on a basic CRUD app, I built:

✅ Full authentication system with JWT
✅ Task management with priority levels and status tracking
✅ Pomodoro timer with session history
✅ Goal tracking with milestones and progress visualization
✅ Bookmark management for quick links
✅ Weekly progress tracking with moon phases
✅ 3D interactive cemetery view
✅ Fully accessible keyboard navigation
✅ Responsive design with spooky animations
✅ Property-based tests for core functionality

The "Aha!" Moment

The moment I realized Kiro had changed my workflow was when I needed to add task statistics to the Settings page.

I just described what I wanted, and within minutes:

The calculations were implemented
The UI was styled consistently with the rest of the app
The responsive grid layout worked perfectly
The animated completion bar had a shimmer effect

It felt less like coding and more like designing the solution, then watching it come to life.

Tips for Using Kiro Effectively

Start with specs - Even for small features, write it down first
Use steering files - Document your tech stack and patterns
Iterate in small chunks - Build feature by feature, not all at once
Trust the context - Kiro remembers your codebase, use that
Refactor confidently - The safety net is real

What's Next for Graveyard Shift?

The app is live and functional, but there's always more to build:

Team collaboration features (shared graveyards)
Mobile app with React Native
Dark ritual animations for completing goals
Integration with calendar apps
Gamification with achievement badges

And with Kiro, I'm actually excited to build these features instead of dreading the complexity.

Try It Yourself

If you're curious about the project:

GitHub: https://github.com/sridurgeshv/Kiroween-Challenge
Tech Stack: React, Node.js, MongoDB, Appwrite

And if you want to experience the Kiro workflow yourself, check out kiro.dev.

Final Thoughts

Building Graveyard Shift taught me that the right tools don't just make you faster - they make you better. Kiro didn't just help me write code; it helped me think through problems, maintain consistency, and ship with confidence.

The future of development isn't about replacing developers with AI. It's about augmenting our abilities, removing friction, and letting us focus on the creative parts of building.

That's what Kiro does. And honestly? I'm hooked.

Built with Kiro for the Kiro Hackathon. #kiro

What's your biggest development workflow pain point? Drop a comment below! 👇

Building Graveyard Shift: How Kiro Changed My Development Workflow

Ritika Srivastava — Thu, 04 Dec 2025 08:11:09 +0000

Building Graveyard Shift: How Kiro Changed My Development Workflow

The Idea: Spooky Productivity

But here's the thing - I wanted to build it fast without sacrificing quality. That's where Kiro came in.

The Old Way vs. The Kiro Way

Before Kiro

My typical workflow looked like this:

Start coding immediately
Realize I forgot something important
Refactor everything
Break something else
Spend hours debugging
Repeat

With Kiro

The workflow became:

Spec it out - Define requirements, design, and tasks
Build iteratively - Implement feature by feature
Get instant feedback - Kiro catches issues immediately
Refactor fearlessly - Context awareness means no broken connections
Ship confidently - Everything works together

The Spec-Driven Approach

One of Kiro's killer features is the spec system. Instead of diving into code, I started by creating structured specifications:

Requirements (EARS Syntax)

WHEN a user completes a task
THEN the system SHALL update the task status to 'completed'
AND the system SHALL record the completion timestamp
AND the system SHALL update the weekly progress moon phase

This forced me to think through the entire feature before writing a single line of code. Game changer.

Design (Property-Based Testing)

Property: Task Status Consistency
GIVEN any task in the system
WHEN the task status is updated
THEN the status MUST be one of: pending, completed, or overdue
AND the status MUST be reflected consistently across all views

By defining properties upfront, I knew exactly what needed to be tested and what "correct" looked like.

Real Examples: Features That Came Together Fast

1. Pomodoro Timer Integration

I wanted users to track focus sessions alongside their tasks. With Kiro:

Defined the spec in 10 minutes
Implemented the backend routes with instant validation
Built the React component with real-time feedback
Integrated with the existing task system seamlessly

Total time: ~2 hours for a fully working feature with backend persistence.

2. 3D Tombstone Cemetery

The visual centerpiece of the app - interactive 3D tombstones that users can click to view task details.

The challenge? Making it accessible, performant, and gracefully degrading for older browsers.

Kiro helped me:

Implement keyboard navigation (arrow keys, Enter, Escape)
Add ARIA labels for screen readers
Detect 3D transform support and fallback gracefully
Handle focus management for modals

All while maintaining the spooky aesthetic.

3. Cursed Dolls (Goal Tracking)

This was the most complex feature - long-term goals with progress tracking, milestone management, and visual feedback.

What would have taken days took hours because:

Kiro understood the entire codebase context
Suggested consistent patterns from existing features
Caught edge cases I would have missed
Helped refactor without breaking anything

The Technical Stack

Frontend:

React 19 with React Router
Framer Motion for animations
Three.js for 3D effects
Tailwind CSS + custom spooky styling

Backend:

Node.js + Express
MongoDB with Mongoose
JWT authentication
Appwrite for user management

Lessons Learned

1. Specs Save Time

Writing specs felt slow at first, but it saved hours of refactoring later. Knowing exactly what you're building before you build it is powerful.

2. Context Awareness is Everything

Not having to explain my codebase repeatedly was liberating. Kiro just knew what I was working with.

3. Iteration Speed Matters

The faster you can try ideas, the better your final product. Kiro's instant feedback loop meant I could experiment without fear.

4. Quality and Speed Aren't Opposites

I shipped faster and with better code quality. Property-based testing, accessibility compliance, and error handling all came naturally.

The Results

In less time than I'd normally spend on a basic CRUD app, I built:

The "Aha!" Moment

The moment I realized Kiro had changed my workflow was when I needed to add task statistics to the Settings page.

I just described what I wanted, and within minutes:

The calculations were implemented
The UI was styled consistently with the rest of the app
The responsive grid layout worked perfectly
The animated completion bar had a shimmer effect

It felt less like coding and more like designing the solution, then watching it come to life.

Tips for Using Kiro Effectively

Start with specs - Even for small features, write it down first
Use steering files - Document your tech stack and patterns
Iterate in small chunks - Build feature by feature, not all at once
Trust the context - Kiro remembers your codebase, use that
Refactor confidently - The safety net is real

What's Next for Graveyard Shift?

The app is live and functional, but there's always more to build:

Team collaboration features (shared graveyards)
Mobile app with React Native
Dark ritual animations for completing goals
Integration with calendar apps
Gamification with achievement badges

And with Kiro, I'm actually excited to build these features instead of dreading the complexity.

Try It Yourself

If you're curious about the project:

GitHub: https://github.com/sridurgeshv/Kiroween-Challenge
Tech Stack: React, Node.js, MongoDB, Appwrite

And if you want to experience the Kiro workflow yourself, check out kiro.dev.

Final Thoughts

The future of development isn't about replacing developers with AI. It's about augmenting our abilities, removing friction, and letting us focus on the creative parts of building.

That's what Kiro does. And honestly? I'm hooked.

Built with Kiro for the Kiro Hackathon. #kiro

What's your biggest development workflow pain point? Drop a comment below! 👇