←back to thread

How to use Claude Code subagents to parallelize development

(zachwills.net)
280 points zachwills | 2 comments | 09 Sep 25 13:21 UTC | HN request time: 0.523s | source
Show context
skimojoe ◴[13 Sep 25 11:20 UTC] No.45231193[source]▶
I am sceptical if these persona based agents really make that much of a difference, and more "appear" to make a difference because of their talk style.

Underneath is just a system prompt, or more likely a prompt layered on top "You are a frontend engineer, competent in react and Next.js, tailwind-css" - the stack details and project layout, key information is already in the CLAUDE.md. For more stuff the model is going to call file-read tools etc.

I think its more theatre then utilty.

What I have taken to doing is having a parent folder and then frontend/ backend/ infra/ etc as children.

parent/CLAUDE.md frontend/CLAUDE.md backend/CLAUDE.md

The parent/CLAUDE.md provides a highlevel view of the stack "FastAPI backend with postgres, Next.js frontend using with tailwind, etc". The parent/CLAUDE.md also points to the childrens CLAUDE.md's which have more granular information.

I then just spawn a claude in the parent folder, set up plan mode, go back and forth on a design and then have it dump out to markdown to RFC/ and after that go to work. I find it does really well then as all changes it makes are made with a context of the other service.

replies(7): >>45231217 #>>45233133 #>>45234479 #>>45235090 #>>45237422 #>>45237827 #>>45238553 #
johntash ◴[13 Sep 25 20:33 UTC] No.45235090[source]▶
I'm also skeptical partially because I don't like the huge essays generated by any llm. CLAUDE.md/AGENTS.md/README.md that are 5+ pages long are all equally bad imo. I prefer following the idea that if something is too verbose for me to want to get anything useful out of it, then the llm should behave similarly. Even if it's not true, why waste 2 paragraphs explaining something that could be explained in one short sentence?

My CLAUDE.md or AGENTS.md is usually just a bulleted list of reminders with high level information. If the agent needs more steering, I add more reminders. I try not to give it _too_ broad of a task without prior planning or it'll just go off the rails.

Something I haven't really experimented with is having claude generate ADRs [1] like your RFC/ idea. I'll probably try that and see how it goes.

[1]: https://adr.github.io/

replies(1): >>45236341 #
1. jrecyclebin ◴[14 Sep 25 00:03 UTC] No.45236341[source]▶
One of the main things I put in my instructions is "hey I'm a solo dev and it's just you and me working on this stuff, so I'm looking for all responses to be concise." I think it helps to give your situation so that the output can be in the proximity of "solo dev" content - which is going to be more concise and practical by nature.

Kind of like telling it to generate Ghibli pics. These things are best at imitation.

replies(1): >>45237845 #
2. johntash ◴[14 Sep 25 06:15 UTC] No.45237845[source]▶
That's probably a good one to keep in there, I'll try it too. Maybe it'll help with the "Here's the plan:" where it starts listing phases and hours/days/weeks for each phase even though it's going to be doing most of the work asap