T-Error

• GIF

  Walkthrough of the t-error frontend

Inspiration

The inspiration for T-Error came from the common frustration that tech leads and developers face when debugging problems. Errors can occur frequently, but understanding their patterns and seeing what is really holding your team up can be tough. We wanted to create something that captures these errors in real time, visualizes them, and lets you write and seamlessly integrate documentation making it easier for teams to build faster.

What it does

T-Error is a terminal error-monitoring tool that captures and logs errors as developers run commands. It aggregates error data in real-time from various client terminals and provides a frontend dashboard to visualize error frequencies and insights, as well as adding the option to seamlessly add documentation. A feature we are really excited about is the ability to automatically run the commands in the documentation without needing to leave the terminal.

How we built it

We built T-Error using: Custom shell: we implemented a custom shell in c++ to capture stderr and seamlessly interface with our backend. Backend: Powered by Node.js, the server collects, processes, and stores error data in mongoDB. Frontend: Developed with React.js, the dashboard visualizes error trends with interactive charts, graphs, and logs, as well as an embedded markdown editor:).

Challenges we ran into

One of the main challenges was ensuring the terminal wrappers were lightweight and didn’t disrupt normal command execution while effectively capturing errors. We spent hours trying to get bash scripts to do what we wanted, until we gave up and tried implementing a shell which worked much better. Additionally, coming up with the UX for how to best deliver existing documentation was a challenge but after some attempts, we arrived at a solution we were happy with.

Accomplishments that we're proud of

We’re proud of building a fully functional MVP that successfully captures and visualizes error data in real-time. Our terminal wrappers integrate seamlessly with existing workflows, and the error analysis and automatic documentation execution has the potential to significantly speed up development.

What we learned

Throughout this project, we learned about the complexities of error logging across multiple environments and how to efficiently process large volumes of real-time data. We also gained experience with the integration of frontend and backend technologies, as well as diving into the lower layers of the tech stack and smoothly chaining everything together.

What's next for T-Error

Going forward, there are a few features that we want to implement. First is error reproduction - we could potentially gain more context about the error from the file system and previous commands and use that context to help replicate errors automatically. We also wanted to automate the process of solving these errors - as helpful as it is to have engineers write documentation, there is a reason there are gaps. This could be done using an intelligent agent for simple tasks, and more complex systems for others. We also want to be able to accommodate better to teams, allowing them to have groups where internal errors are tracked.

Built With

• c++
• javascript
• mongodb
• nextjs
• node.js
• react.js
• typescript