GitStory is a full-stack web app that transforms any public GitHub profile into a beautiful, AI-generated career narrative page that can be shared in seconds. The idea is simple: your GitHub already contains the shape of your builder journey, but most people never turn that raw activity into a story. GitStory does that for you.
Enter a GitHub username, and GitStory fetches real public profile data: repositories, stars, followers, languages, first repo, top projects, and contribution signals. It then sends that grounded profile context to GPT-5.4, OpenAI's current frontier model for professional work, through Replit AI Integrations. The result is not a generic bio. It is a cinematic developer narrative with chapters, a superpower quote, fun stats, and a legacy statement that captures who you are as a builder.
Live project
git-story-maker--sairammaruri.replit.appWhy I Built It
GitHub profiles are packed with signal, but they usually read like dashboards: repo counts, language bars, stars, followers, and pinned projects. Those metrics matter, but they rarely explain the person behind the work. GitStory turns those public signals into something more human: a page that feels like a career trailer for a developer.
For Replit's 10-year celebration Buildathon at buildathon.replit.com, I wanted to build something that matched the spirit of Replit itself: fast creation, public sharing, and AI as a collaborator that helps people express what they have built. GitStory takes a profile that already exists and gives it a shareable narrative layer.
What GitStory Does
- Pulls live GitHub data for any public profile.
- Generates a unique career story using GPT-5.4, grounded in real repos and profile history.
- Displays the output as a full-page cinematic story at
/story/:username. - Caches generated stories in PostgreSQL so repeat visits are instant.
- Lets users share their story on X with one click or copy the link directly from the nav bar.
- Includes an Explore page showing all stories generated so far.
- Shows a live animated counter for total stories generated on the home page.
- Rate limits story generation to 3 requests per IP every 5 minutes to prevent abuse.
The Product Experience
The home page starts with a GitHub-inspired dark interface and a direct input: type a username, generate a story, and get a shareable page. Once generated, the story is structured like a narrative profile rather than a stats card. It has chapters, personality, momentum, and a clear sense of what makes that developer's work distinctive.
Every story gets a permanent route, so the output can be shared like a portfolio page. That was important to the design: the app is not just a private generator. It creates public artifacts people can send to friends, recruiters, teammates, or the Buildathon community.
Tech Stack
- Frontend: React and Vite.
- Backend: Express.
- Database: PostgreSQL with Drizzle ORM.
- AI generation: GPT-5.4 through the Replit AI Integrations proxy.
- API workflow: OpenAPI-first contract with auto-generated React Query hooks.
- Design: dark GitHub-inspired theme using
#0d1117with Replit orange accents using#f97316. - Build process: built entirely with Replit Agent for Replit's 10-year celebration Buildathon.
Why the AI Layer Matters
The AI layer is not there to invent a fake identity. It is there to translate structured public activity into a story people can actually read. GitStory gives the model grounded facts from GitHub first, then asks it to shape those facts into a memorable narrative: where the builder started, what patterns appear in their work, what technologies they lean toward, and what their public projects suggest about their developer identity.
That grounding is the heart of the app. A good generated story should feel personal because it is attached to real public work, not because it is decorated with vague motivational language.
The Core Idea
Your GitHub has a story. Most people never tell it. GitStory tells it for you, powered by GPT-5.4 and built on Replit.
GitStory is my submission for Replit's 10-year celebration Buildathon at buildathon.replit.com. It is a small but complete product: live data, generated stories, persistent caching, shareable routes, an Explore page, abuse protection, and a visual identity built around GitHub's developer culture and Replit's orange energy.
Try it here:
Generate your GitStorySources & References
GitStory was a fun build that packed a full product into a buildathon timeline. If you are experimenting with AI-native apps, the fastest way to learn is to ship one end to end.
Have thoughts or want to discuss further? Connect with me on LinkedIn or explore more on my portfolio.
Until next time,
Sairam Maruri
Note: This article includes AI-assisted writing and manual curation. Use at your own risk.