๐Ÿ’ก Inspiration

Every developer wants seamless deployment, but setting up CI/CD pipelines can be confusing and time-consuming โ€” especially for beginners. We imagined a world where deploying a project is as easy as running one command:

npx deployated <repo-url>

##๐Ÿš€ What it does
Deployated is a CLI tool that:

Auto-detects your projectโ€™s tech stack (Node.js, Flask, Spring Boot)

Uses the Gemini API to generate a smart CI/CD workflow

Pushes the workflow back to your GitHub repo

Triggers deployment to platforms like Vercel or Render

##๐Ÿ›  How we built it
We split the project across 3 key roles:

๐Ÿง  AI Agent: Prompted Gemini using repo files like package.json or requirements.txt

๐Ÿ” GitHub Integrator: Used GitHubโ€™s REST API to push workflows into repos securely

๐Ÿš€ CLI Engine: Created a smooth npx CLI that detects โ†’ generates โ†’ deploys in one flow

We used Node.js, commander.js, axios, and Gemini for AI-powered workflow generation.

##๐Ÿง—โ€โ™€๏ธ Challenges we ran into
Gemini API sometimes returned partial or invalid YAML โ€” we added fallback templates as backups

Mapping diverse repo structures to specific tech stacks wasnโ€™t always obvious

GitHub token handling and workflow injection required careful error management

##๐Ÿ… Accomplishments that we're proud of
Building a fully working MVP in just a few days

Plug-and-play CLI that actually deploys real projects

Clean separation of concerns between AI, GitHub, and CLI logic

##๐Ÿ“š What we learned
How to craft effective prompts for CI/CD pipelines using AI

The structure and quirks of GitHub Actions workflows

How to package and publish a CLI with npx functionality

##๐Ÿ”ฎ What's next for Deployated
 Support more frameworks like Django

 Add rollback support for failed deploys

 Integrate with MongoDB to store deployment logs

 Make Gemini-generated workflows editable in a web UI

Built With

Share this project:

Updates