@LAZX

I built Chronicle, a local-first developer intelligence tool that scans a project’s history and changes, stores them as observations, analyzes those observations into findings, and then uses an AI interpreter to answer questions about how and why things changed in the project. The most challenging part for me was state management, especially keeping track of which observations had already been analyzed so Chronicle could process only new changes instead of repeatedly analyzing everything. I had to think carefully about scan state and analysis state, checkpoints, and how different scanners such as Git and Django migrations could work together without duplicating or losing information. I’m most proud of how Chronicle evolved into a proper pipeline rather than just a collection of commands: scanners produce observations, analyzers produce findings, and the interpretation layer connects those findings back to relevant project history to answer a developer’s questions. For someone testing the project, they should initialize Chronicle in a project with chronicle init, make sure the required AI provider and model are configured along with an API key, and then create some Git commits or Django migrations and run the appropriate scan, analyze, and interpret commands. The best way to test it is to make a change, scan it, analyze it, and then ask Chronicle a question about that change to see whether it can trace the history and explain what happened and why.

I made a simple portfolio like page according to the guide as I completely vibecoded this I didn’t have any challenges. I have not checked the page properly as its just a simple page showing basic info.