Agent best practices - grind mode and various hooks
A comprehensive guide to working with coding agents, from starting with plans to managing context, customizing workflows, and reviewing code.
Summary
LLM coding agents are ineffective at basic code manipulation tasks such as copy-paste, instead relying on rewriting code from memory, which leads to awkward and less reliable refactoring.
They also fail to ask clarifying questions before making major changes, preferring to proceed with assumptions and brute-force solutions. These limitations make their workflow feel unnatural compared to human developers and hinder effective collaboration.