Inspiration: I was inspired to create NetDiag after noticing how many users panic when a webpage fails to load and are met with cryptic browser errors like ERR_CONNECTION_RESET. I wanted to make network errors understandable — turning frustration into clarity.
How I Built It: The project is a Chrome extension built with HTML, JavaScript, and the Chrome Extensions API. It listens for network request failures via chrome.webRequest and displays a custom popup with friendly messages and quick fixes. For offline AI help, I integrated a locally running Ollama model (e.g., phi3:mini), which generates short, context-based repair tips using natural language.
What I Learned: I learned how browser networking works under the hood — from DNS resolution to SSL handshakes — and how to interface with local AI models without relying on external APIs. I also deepened my understanding of Chrome’s extension sandboxing and messaging system.
Challenges: The main challenge was handling offline inference efficiently. Running a local model without freezing the UI required asynchronous calls and a graceful fallback when Ollama wasn’t available. Another challenge was crafting explanations that are short, helpful, and understandable to non-technical users.
Result: NetDiag empowers users to troubleshoot connectivity issues confidently — even without internet or technical knowledge.
Log in or sign up for Devpost to join the conversation.