Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How to use Claude Code subagents to parallelize development

zachwills.net

Post reply on HN
Print view View original

21–30 of 131 posts

Re: How to use Claude Code subagents to parallelize development

#21

Post by NitpickLawyer »

mattmanser wrote: post #10
olivermuty wrote: post #9

Earlier quoted context omitted.

This is only a problem if an agent is made in a lazy way (all of them). Chat completion sends the full prompt history on every call. I am working on my own coding agent and seeing massive improvements by rewriting history using either a smaller model or a freestanding call to the main one. It really mitigates context poisoning.

Everyone complains that when you compact the context, Claude tends to get stupid Which as far as I understand it is summarizing the context with a smaller model. Am I misunderstanding you, as the practical experience of most people seem to contradict your results.

One key insight I have from having worked on this from the early stages of LLMs (before chatgpt came out) is that the current crop of LLM clients or "agentic clients" don't log/write/keep track of success over time. It's more of a "shoot and forget" environment right now, and that's why a lot of people are getting vastly different results. Hell, even week to week on the same tasks you get different results (see the recent claude getting dumber drama).

Once we start to see that kind of self feedback going in next iterations (w/ possible training runs between sessions, "dreaming" stage from og RL, distilling a session, grabbing key insights, storing them, surfacing them at next inference, etc) then we'll see true progress in this space.

The problem is that a lot of people work on these things in silos. The industry is much more geared towards quick returns now, having to show something now, rather than building strong fo0undations based on real data. Kind of an analogy to early linux dev. We need our own Linus, it would seem :)

Re: How to use Claude Code subagents to parallelize development

#23

Post by troupo »

NitpickLawyer wrote: post #21
mattmanser wrote: post #10

Earlier quoted context omitted.

Everyone complains that when you compact the context, Claude tends to get stupid Which as far as I understand it is summarizing the context with a smaller model. Am I misunderstanding you, as the practical experience of most people seem to contradict your results.

One key insight I have from having worked on this from the early stages of LLMs (before chatgpt came out) is that the current crop of LLM clients or "agentic clients" don't log/write/keep track of success over time. It's more of a "shoot and forget" environment right now, and that's why a lot of people are getting vastly different results. Hell, even week to week on the same tasks you get different results (see the r…

> don't log/write/keep track of success over time.

How do you define success of a model's run?

Re: How to use Claude Code subagents to parallelize development

#24

Post by ako »

NitpickLawyer wrote: post #21
mattmanser wrote: post #10

Earlier quoted context omitted.

Everyone complains that when you compact the context, Claude tends to get stupid Which as far as I understand it is summarizing the context with a smaller model. Am I misunderstanding you, as the practical experience of most people seem to contradict your results.

One key insight I have from having worked on this from the early stages of LLMs (before chatgpt came out) is that the current crop of LLM clients or "agentic clients" don't log/write/keep track of success over time. It's more of a "shoot and forget" environment right now, and that's why a lot of people are getting vastly different results. Hell, even week to week on the same tasks you get different results (see the r…

I’ve experimented with feature chats, so start a new chat for every change, just like a feature branch. At the end of a chat I’ll have it summarize the the feature chat and save it as a markdown document in the project, so the knowledge is still available for next chats. Seems to work well.

You can also ask the llm at the end of a feature chat to prepare a prompt to start the next feature chat so it can determine what knowledge is important to communicate to the next feature chat.

Summarizing a chat also helps getting rid of wrong info, as you’ll often trial and error towards the right solution. You don’t want these incorrect approaches to leak into the context of the next feature chat, maybe just add the “don’t dos” into a guidelines and rules document so it will avoid it in the future.

Re: How to use Claude Code subagents to parallelize development

#25

Post by NitpickLawyer »

troupo wrote: post #23
NitpickLawyer wrote: post #21

Earlier quoted context omitted.

One key insight I have from having worked on this from the early stages of LLMs (before chatgpt came out) is that the current crop of LLM clients or "agentic clients" don't log/write/keep track of success over time. It's more of a "shoot and forget" environment right now, and that's why a lot of people are getting vastly different results. Hell, even week to week on the same tasks you get different results (see the r…

> don't log/write/keep track of success over time. How do you define success of a model's run?

Lots of ways. You could do binary thumbs up/down. You could do a feedback session. You could look at signals like "acceptance rate" (for a pr?) or "how many feedback messages did the user send in this session", and so on.

My point was more on tracking these signals over time. And using them to improve the client, not just the model (most model providers probably track this already).

Re: How to use Claude Code subagents to parallelize development

#26

Post by sixhobbits »

I often see people making these sub agents modelled on roles like product manager, back end developer, etc.

I spent a few hours trying stuff like this and the results were pretty bad compared to just using CC with no agent specific instructions.

Maybe I needed to push through and find a combination that works but I don't find this article convincing as the author basically says "it works" without showing examples or comparing doing the same project with and without subagents.

Anyone got anything more convincing to suggest it's worth me putting more time into building out flows like this instead of just using a generic agent for everything?

Re: How to use Claude Code subagents to parallelize development

#27

Post by ixsploit »

olivermuty wrote: post #9
theshrike79 wrote: post #7

Earlier quoted context omitted.

I don't use subagents to do things, they're best for analysing things. Like "evaluate the test coverage" or "check if the project follows the style guide". This way the "main" context only gets the report and doesn't waste space on massive test outputs or reading multiple files.

This is only a problem if an agent is made in a lazy way (all of them). Chat completion sends the full prompt history on every call. I am working on my own coding agent and seeing massive improvements by rewriting history using either a smaller model or a freestanding call to the main one. It really mitigates context poisoning.

I do something similar and I have the best results of not having a history at all, but setting the context new with every invokation.

Re: How to use Claude Code subagents to parallelize development

#28

Post by redrove »

sixhobbits wrote: post #26

I often see people making these sub agents modelled on roles like product manager, back end developer, etc. I spent a few hours trying stuff like this and the results were pretty bad compared to just using CC with no agent specific instructions. Maybe I needed to push through and find a combination that works but I don't find this article convincing as the author basically says "it works" without showing examples or…

This has been my experience so far as well. It seems like just basic prompting gets me much further than all these complicated extras.

At some point you gotta stop and wonder if you’re doing way too much work managing claude rather than your business problem.

Re: How to use Claude Code subagents to parallelize development

#29

Post by cpursley »

sixhobbits wrote: post #26

I often see people making these sub agents modelled on roles like product manager, back end developer, etc. I spent a few hours trying stuff like this and the results were pretty bad compared to just using CC with no agent specific instructions. Maybe I needed to push through and find a combination that works but I don't find this article convincing as the author basically says "it works" without showing examples or…

I think the trick is the synthesize step which brings the agents findings together. That's where I've had the most success, at least.

Re: How to use Claude Code subagents to parallelize development

#30

Post by posix86 »

simianwords wrote: post #12

Slightly off topic but I would really like agentic workflow that is embedded in my IDE as well as my code host provider like GitHub for pull requests. Ideally I would like to spin off multiple agents to solve multiple bugs or features. The agents have to use the ci in GitHub to get feedback on tests. And I would like to view it on IDE because I like the ability to understand code by jumping through definitions. Suppo…

This already exists. Look at cursor with Linear, you can just reply with @cursor & some instructions and it starts working in a vm. You can watch it work on cursor.com/agents or using the cursor editor. Result is a PR. Also github has copilot getting integrated in the github ui, but not that great in my experience
Post reply on HN