Code-Weaver

• 

  System-Core

• 

  Git-Lab Integration

• 

  Features

Core Application is a TkInter application (Provided in Github) The Application is made using Google Cloud + Gemini + Gitlab and Hosted Using Google Cloud Run

Inspiration

As a solo developer constantly switching between writing code, debugging, researching solutions, and setting up CI/CD pipelines, I found that the biggest productivity killer wasn’t complexity—it was context-switching. I wanted to build a platform that eliminates the mental overhead of development by creating a true Co-Coding experience, where AI is not just an add-on, but a real-time collaborator.

The GitLab + Google Cloud Hackathon challenge was the perfect catalyst. It inspired me to combine AI, cloud analytics, and version control into a unified, intelligent development environment: CodeWeaver AI.

What it does

CodeWeaver AI is a desktop-based, AI-powered DevSecOps platform that redefines how solo developers and teams build software. It provides:

• AI Co-Coding Chat using Google Gemini for real-time, context-aware code assistance
• One-click debugging with Gemini-powered "Fix Error" feature
• Instant Stack Overflow search via BigQuery with automated query generation
• Actionable code quality analysis with direct-edit suggestions
• AI-generated .gitlab-ci.yml pipeline files, customized for the user’s project
• Integrated GitLab push/commit, closing the DevSecOps loop within the same interface As a solo builder, I structured the entire app using the following stack:

Frontend & Desktop App: Built using Python and Tkinter for the UI, designed to remain lightweight and fast

AI Integration: Powered by Google Gemini, used for debugging, refactoring, natural language coding, and CI/CD generation

Google-Cloud BigQuery Integration: Used the bigquery-public-data.stackoverflow dataset to enable direct querying and smart search through billions of Stack Overflow posts

GitLab Integration: Used the GitLab API for seamless commit, push, and CI/CD integration

Architecture: Designed for simplicity and speed, with serverless query execution on Google Cloud, and context-passing pipelines for AI reasoning It’s built to help developers write better code, fix problems faster, and deploy seamlessly—without ever leaving the app.

How we built it

As a solo developer, I built CodeWeaver AI from the ground up with a modular architecture focused on speed, reliability, and intelligence. The platform integrates deeply with three major components:

• Google Gemini: Acts as the cognitive engine. It powers everything from debugging to code analysis, CI/CD generation, and conversational coding. It’s used through carefully structured prompts that feed in terminal context, code state, and user queries to produce actionable results.
• Google Cloud BigQuery: Handles querying of the massive bigquery-public-data.stackoverflow dataset. I implemented dynamic query generation so developers can get instant, relevant solutions to their coding problems—either manually or through Gemini’s automated queries.
• GitLab Integration: The entire DevSecOps flow is integrated using GitLab’s API. CodeWeaver can generate .gitlab-ci.yml files with AI, push them directly to a GitLab repository, and manage commits—all from within the app. ## Challenges we ran into Tkinter Applications can not be deployed on web to obtain an URL which is a must for this hackathon, and I had to redevelop the entire solution from scratch. ## Accomplishments that we're proud of Being able to redo and pull off the entire project in the last few hours of submission ## What we learned To recognize REAL problems and to solve them with the power of tech and innovation ## What's next for Code-Weaver To integrate CI/CD pipeline, add directory building assisted with GEMINI to save time and effort. UI upgrades

Built With

• python
• react
• tkinter