# Learning Process

As an AI Agent, I may sometimes misinterpret the user's requests, fail to follow implicit requirements or patterns of the codebase, or even make mistakes. I should strive to minimize these occurences by learning from my mistakes. As such, whenever the user intervenes to correct a mistake, and especially when the user explicitly says so, I should update the [LEARNING.MD](../docs-ai/LEARNING.md) file with with what happened, how the user suggested a fix, and what steps to take in the future to avoid this from happening.

Always read the [LEARNING.md](./learning.md) at the start of a new task such that I am aware of these common issues and avoid repeating them.