Times are changing. I have barely written a line of code by hand throughout January. Most of my time was spent telling Claude and Codex what to do, plan out what to do or have them fix their own mess and writing rules about that.
Here are the three things, that half the internet talks about, but I think are conceptually relevant to make AI programming go brrr regardless of the LLM provider you use (OpenCode, Claude Code, Codex, what have you).
Plan mode
The first and least important is to use plan mode. It’s what we have done since the dawn of LLMs. It also burns a few tokens most of the time, but since there typically is a written document by the end of it, compaction and context management is not that important anymore.
Patterns
Patterns are like really, really important. If you want an LLM (especially Claude) to be maximally efficient in implementing something, reference a similar implementation in the codebase and the agent will fly.
I had that in the back of my mind all the time, but I just recently really got it, when I want Claude to migrate some code from an old library and its usages to a new version. The first time I tried to just let it rip. That didn’t really work. But then I converted one of the 96 files by hand and told it to reference that manually converted file and it one shotted the rest.
So write down your patterns, reference them across the codebase and what the agent produces will probably be much better, then when you let it run free.
Feedback loops
The most important part is (automated) feedback loops. I think the internet has largely covered it, but I cannot overstate how important that is. Find a way for your coding agent to check its own work. Of course that’s not possible in all cases, but where it is, it’s wild.
Here are some ideas:
- https://ampcode.com/notes/feedback-loopable
- https://x.com/i/status/2010719790659408260
- Write your own script, that can debug what you want, then let your agent use it.
- Write an agent who’s sole purpose is to play devils advocate and check against a list of patterns, you tell it to.
DSLs
Sometimes it’s also a really good idea to implement a small, composable DSL to constrain the output of an LLM. For example, if you have a DSL to build user defined forms, you can tell the LLM that and for example ask it to build a welcome form and it will do it. Same for drawing stuff for example.
type Widget
= { type: 'Paragraph', content: Text[] }
| { type: 'Input', name: string, placeholder: string }
| { type: 'Row', name: string, placeholder: string };
type Text = {
content: string;
options: FormattingOptions;
}What a time to be alive.