Git --Profile Print

Inspiration

I built Git --Profile Print because most learning roadmaps ignore what a developer has already created. I wanted to use real GitHub work to answer a simple question: What should I learn and prove next?

What it does

Git --Profile Print analyzes a public GitHub profile, maps skills to repository evidence, identifies important gaps, and creates a personalized four-week learning sprint. Every week ends with visible proof, such as a test, benchmark, evaluation dataset, commit, or deployment.

How I built it

I built the application using Next.js, TypeScript, the GitHub API, OpenAI GPT-5.6, and Codex. A deterministic engine scores repository evidence, while GPT-5.6 can turn that evidence into structured recommendations. I also added progress tracking, downloadable sprint briefs, and ready-to-use Codex prompts.

Challenges and lessons

My biggest challenge was making the recommendations transparent instead of creating another generic AI roadmap. I solved this by connecting every skill signal to real repositories.

I also faced GitHub rate limits and deployment issues with Next.js runtime chunks. I added a reliable fallback mode and changed the production build process to make the application stable.

I learned that AI learning advice becomes more useful when users can inspect the evidence and finish with work that others can verify.

Built With

• api
• codex
• github
• gpt-5.6
• next.js
• openai-gpt-5.6
• typescript