byme8

You Don't Need a CLAUDE.md

In this article, I want to discuss how to better use LLM agents, specifically how to manage the CLAUDE.md/AGENTS.MD file effectively and it seems like there is no point it in.

Before we start

This article will not cover the standard prompt engineering techniques that have been described many times by Anthropic and other LLM providers. You won't find advice here like "avoid negative examples," "set the LLM role," "be specific in your instructions," or similar well-documented practices. If you need that foundation, check the official documentation from Anthropic or OpenAI—they cover it thoroughly.

Instead, this article focuses on how my workflow evolved through months of using Claude Code on real projects. It's about the practical patterns I discovered for organizing documentation and keeping the LLM effective as projects grow in complexity.

byme8

Experienced Team

Experienced professionals never proceed directly to completion, but follow from the starting point along an arc, consisting of points equidistant from both the starting point and the result point (circumference).

After traveling halfway, the team usually stops to think and rushes toward completing the task. (fig. 1) Thus, the formula for expected execution time looks like this:

t = R × (π + 1) + β

where t is the time spent to achieve the goal, and R is correctly estimated initial timeframes for project completion.

byme8

DuckInterface

I was playing with new .Net 5 and the Source Generator lately and got an idea that it is possible to add "duck typing" support to C#. I would say it is purely academic(no one will use it in production I hope), but it is fun stuff so I decided to try.

The nuget package with results you can find here

The repository is here: https://github.com/byme8/DuckInterface