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

Claude Managed Agents

claude.com

Post reply on HN
Print view View original

91–100 of 109 posts

Re: Claude Managed Agents

#91

Post by suncemoje »

gck1 wrote: post #85
jameslk wrote: post #27

We're in the early days of agentic frameworks, like the pre-PHP web. CGI scripts and webmasters. Eventually the state-of-the-art will slow down and we'll eventually have something elegant like Rails come out. Until then, every agent framework is completely reinvented every week due to new patterns and new models. evals, ReACT, DSPy, RLM, memory patterns, claws, dynamic context, sandbox strategies. It seems like locki…

I believe framework is simply never, ever going to work for LLM-based agentic workflows. Framework is simply way too rigid for a non-deterministic technology. We may see libraries that provide tools for managing agents, but then again, there's nothing that tmux can't do already.

I've been using the OpenAI Agents SDK for a while now and am largely happy with the abstractions - handoffs/sub-agents, tools, guardrails, structured output, etc. Building the infra and observability, and having it scale reliably, was a bigger pain for me. So I do get Anthropic's move into managed agents.

Re: Claude Managed Agents

#92

Post by suncemoje »

I am wondering if this is (or what else will be) the last piece of software + infra that's needed to "automate it all" and have non-technical people build, run, and maintain it? To me the all this agentic workflow automation is headed that way. Am I missing something?

Re: Claude Managed Agents

#93

Post by jguetzkow »

Interesting that the entire discussion here is about orchestration, vendor lock-in, and model selection — but nobody is asking about the output. These agents run for hours, write code across multiple repos, and open PRs autonomously. Anthropic built solid infrastructure governance: sandboxing, scoped permissions, execution tracing. That's the "can the agent access this system safely?" question, answered. But there's a different question that nobody seems to be answering: is the generated code actually correct? Not syntactically — it almost always is. I mean semantically. Does it reference database fields that actually exist in the schema? Does it call API routes that are actually defined? Does it handle env variables that are actually set? Does it meet compliance requirements that apply to the system it's modifying? 45% of AI-generated code contains security vulnerabilities (Veracode). Code duplication has quadrupled (GitClear). And we're now scaling this with autonomous agents that run unsupervised for hours. The orchestration problem is getting solved. The governance-of-output problem is wide open. That's the layer that's missing.

Re: Claude Managed Agents

#94

Post by johnphilipp »

suncemoje wrote: post #92

I am wondering if this is (or what else will be) the last piece of software + infra that's needed to "automate it all" and have non-technical people build, run, and maintain it? To me the all this agentic workflow automation is headed that way. Am I missing something?

[dead]

Re: Claude Managed Agents

#95

Post by spwa4 »

gck1 wrote: post #86
spwa4 wrote: post #81

Earlier quoted context omitted.

Well, that sucks. Replacing the harness with something task-specific has proven very powerful in my usecases. But you should correct: Claude is very happy to let you use whatever you want for a harness ... as long as you're on a pay-as-you-go plan. So it's not blocked. It's just not allowed on the $20 per month plan. First, harnesses can give access to company internal tools (like the ticket queue). You could do this…

> A trivial example is that you can improve performance in a very simple way: ask "are you sure?" showing the model what it intends to do, BEFORE doing it. Improves performance by 10% Put it into the "are you sure" loop and you'll see the model will just keep oscillating for eternity. If you ask the model to question the output, it will take it as an instruction that it must, even if the output is correct.

Not in my experience. I mean, it happens. But models can check if their own function calls are reasonable. And that doesn't require dropping the context cache, so it's a lot less expensive than you would probably initially think.

Re: Claude Managed Agents

#96

Post by jawiggins »

Jayakumark wrote: post #83
jawiggins wrote: post #42

Earlier quoted context omitted.

Shameless self promo but, I've been working on Optio specifically for coding, it works by taking any harness you want and tasking it to open Github/lab PRs based on notion/jira/linear tickets, see: https://news.ycombinator.com/item?id=47520220 It works on top of k8s, so you can deploy and run in your own compute cluster. Right now it's focused only on coding tasks but I'm currently working on abstractions so you can…

@jawiggins saw your repo it looks like openAI symphony but better as it works across multiple agents and issue trackers and the feedback loop is great . One feature request though - can you add plan mode ? Your issues are so detailed it becomes plan to implement (but I guess your plan mode is currently happening outside of GitHub issues ) but let’s say issue is “implement support for plan mode” there should be back a…

Thanks for the feedback. Earlier I expected I'd need to do more back and forth with the agents before accepting their work but in general I've found it isn't needed.

I do have some features coming up that will improve the ability to converse with the agent as it's running. I'll make a note to add in a plan setting so you can have that run and converse before it gets going.

Post reply on HN