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

The creator of Claude Code's Claude setup

twitter.com

Post reply on HN
Print view View original

11–20 of 423 posts

Re: The creator of Claude Code's Claude setup

#11

Post by gulaggoon »

andruby wrote: post #7

Great list of useful tips. It's interesting that Boris doesn't mention "Agent Skills" at all. I'm still a bit confused at the difference between slash commands and Agent Skills. https://code.claude.com/docs/en/skills

The main difference is that slash commands are invoked by humans, whereas skills can only be invoked by the agent itself. It works kinda as conditional instructions.

As an example, I have skills that aide in adding more detail to plans/specs, debugging, and for spinning up/partitioning subagents to execute tasks. I don't need to invoke a slash command each time, and the agent can contextually know by the instructions I give it what skills to use.

Re: The creator of Claude Code's Claude setup

#12

Post by codesnik »

Aurornis wrote: post #5
bikeshaving wrote: post #3

Must be nice to have unquota’ed tokens to use with frontier AI (is this the case for Anthropic employees?). One thing I think is fascinating as we enter the Intellicene is the disproportionate access to AI. The ability to petition them to do what you want is currently based on monthly subscriptions, but will it change in the future? Who knows?

> (is this the case for Anthropic employees?) It would be funny if the company paying software engineers $500K or more along with gold-plated stock options was limiting how much they could use the software their company was developing.

well, not only their software but also hardware resources they're renting, but I agree they don't.

Re: The creator of Claude Code's Claude setup

#14

Post by perfmode »

Aurornis wrote: post #5
bikeshaving wrote: post #3

Must be nice to have unquota’ed tokens to use with frontier AI (is this the case for Anthropic employees?). One thing I think is fascinating as we enter the Intellicene is the disproportionate access to AI. The ability to petition them to do what you want is currently based on monthly subscriptions, but will it change in the future? Who knows?

> (is this the case for Anthropic employees?) It would be funny if the company paying software engineers $500K or more along with gold-plated stock options was limiting how much they could use the software their company was developing.

Tokens aren’t free.

Re: The creator of Claude Code's Claude setup

#15

Post by cafebeen »

andruby wrote: post #7

Great list of useful tips. It's interesting that Boris doesn't mention "Agent Skills" at all. I'm still a bit confused at the difference between slash commands and Agent Skills. https://code.claude.com/docs/en/skills

I believe slash commands are all loaded into the initial context and executed when invoked by the user. Skills on the other hand only load the name and description into initial context, and the agent (not user) determines when to invoke them, and only then is the whole skill loaded into context. So skills shift decision making to the agent and use progressive disclosure for context efficiency.

Re: The creator of Claude Code's Claude setup

#16

Post by perfmode »

seg_lol wrote: post #10
joshribakoff wrote: post #6

Earlier quoted context omitted.

Why is that funny? What company gives you unlimited resources? That doesn’t scale. Google employees can’t just demand a $10,000 workstation. It’s reasonable to assume they have some guardrails, for both financial and stability reasons. Who knows… if it’s unlimited now, will it stay that way forever? Probably unlimited in the same sense as unlimited pto.

Folks in MEGACORP cloud env can spend > 5 digits a month and not get noticed.

Can they spend 7 digits?

Re: The creator of Claude Code's Claude setup

#18

Post by baalimago »

I'm a bit jealous. I would like to experiment with having a similar setup, but 10x Opus 4.5 running practically non stop must amount to a very high inference bill. Is it really worth the output?

From experimentation, I need to coach the models quite closely in order to get enough value. Letting it loose only works when I've given very specific instructions. But I'm using Codex and Clai, perhaps Claude code is better.

Re: The creator of Claude Code's Claude setup

#19

Post by dcre »

bikeshaving wrote: post #3

Must be nice to have unquota’ed tokens to use with frontier AI (is this the case for Anthropic employees?). One thing I think is fascinating as we enter the Intellicene is the disproportionate access to AI. The ability to petition them to do what you want is currently based on monthly subscriptions, but will it change in the future? Who knows?

Pretty sure I have seen them imply in one of the panel discussions on their YouTube channel (can't remember which) that they get unlimited use of the best models. I remember them talking about records for the most spent in a day or something.
Post reply on HN