Agentic Coding Agents

We should be using LLM agents to improve the readability of our software, because writing the software was never the hard part. The agent can do a lot when provided the right information. But in the absence of any readable code and a good design, agents can only go so far. The code should be improved such that it can be better understood by both AI and humans. After all, most of the work comes after the code is written.