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

Claude Skills are awesome, maybe a bigger deal than MCP

simonwillison.net

Post reply on HN
Print view View original

271–280 of 383 posts

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#271

Post by Groxx »

Herring wrote: post #219
michael1999 wrote: post #203

We're doing something like this internally. Our monorepo context files were much too big, so we built a progressive tree of fragments to load up for different tasks. I am struck by how much these kinds of context documents resemble normal developer documentation, but actually useful and task-oriented. What was the barrier to creating these documents before? Three theories on why this is so different: 1) The feedback…

Probably all the same reasons tech debt exists in the first place: business pressure, poor design, lack of resources. It used to be expensive to keep good documentation up to date as the code changes.

yep. it's suddenly more-obviously valuable, so it's getting done.

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#272

Post by Yeroc »

simonw wrote: post #37
rco8786 wrote: post #20

Earlier quoted context omitted.

I don’t really see how this replaces MCP tbh. MCP gives the LLM access you your APIs. These skills are just text files with context about how to perform specific tasks.

You don't need MCP if you can instead drop in a skill markdown file that says "to access the GitHub API, use curl against api.github.com and send the GITHUB_API_KEY environment variable in the authorization header. Here are some examples. Consult github-api.md for more."

That's going to be a lot less efficient context-wise and computing-wise than using either a purpose-built MCP or skill based around executing a script.

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#273

Post by tptacek »

skydhash wrote: post #258
tptacek wrote: post #255

Earlier quoted context omitted.

I've been a Usenix reviewer twice, once as a program chair (I think that's what they call the co-leaders of a PC?). So this doesn't clarify anything for me.

To out it more clearly. You take a domain (like OS security, perfomance, and administration) and you’ll find those kinds of problems that people feel important to share solutions with each other. Solutions that are not trivially found. Findings you can be proud your name is attached with. And then you have something like the LLM craze where while it’s new, it’s not improving any part of the problem it’s supposed to s…

OK: I think I have decisively established my Usenix bona fides here, and I'm repeating my original question: what is the cadence at which we resolved "deep question" prior to the era of LLMs? (It began in 2023.)

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#274

Post by visarga »

frankc wrote: post #10

So far I am in the skeptic camp on this. I don't see it adding a lot of value to my current claude code workflow which already includes specialized agents and a custom mcp to search indexed mkdocs sites that effectively cover the kinds of things I would include in these skills file. Maybe it winds up being a simpler, more organized way to do some of this, but I am not particularly excited right now. I also think "ski…

> and a custom mcp to search indexed mkdocs sites that effectively cover the kinds of things I would include in these skills file

Search and this document base pattern are different. In search the model uses a keyword to retrieve results, here the model starts from a map of information, and navigates it. This means it could potentially keep context better, because search tools have issues with information fragmentation and not seeing the big picture.

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#275

Post by hollowturtle »

simonw wrote: post #261
poopiokaka wrote: post #259

[flagged]

I'm not. Here are my disclosures: https://simonwillison.net/about/#disclosures And if my disclosures aren't enough for you, here's the FTC explaining how it would be illegal for an AI vendor to pay someone to write something like this without both sides disclosing the relationship: https://www.ftc.gov/business-guidance/resources/ftcs-endorse...

> I have not accepted payments from LLM vendors, but I am frequently invited to preview new LLM products and features from organizations that include OpenAI, Anthropic, Gemini and Mistral, often under NDA or subject to an embargo. This often also includes free API credits and invitations to events.

You don't need money, just incentives and network effects

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#276

Post by simonw »

hollowturtle wrote: post #275
simonw wrote: post #261

Earlier quoted context omitted.

I'm not. Here are my disclosures: https://simonwillison.net/about/#disclosures And if my disclosures aren't enough for you, here's the FTC explaining how it would be illegal for an AI vendor to pay someone to write something like this without both sides disclosing the relationship: https://www.ftc.gov/business-guidance/resources/ftcs-endorse...

> I have not accepted payments from LLM vendors, but I am frequently invited to preview new LLM products and features from organizations that include OpenAI, Anthropic, Gemini and Mistral, often under NDA or subject to an embargo. This often also includes free API credits and invitations to events. You don't need money, just incentives and network effects

I quote, "Bro is paid to write this".

And yeah, blogging does kind of work on incentives. If I write things and get good conversations as a result, I'm incentivized to write more things. If I write something and get silence then maybe I won't invest as much time in the future.

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#277

Post by visarga »

fny wrote: post #121
mooreds wrote: post #111

Reposting a comment I made when it was posted 10 hours ago: As someone who is looking into MCP right now, I'd love to hear what folks with experience in both of these areas think. My first impressions are that MCP has some advantages: - around for longer and has some momentum - doesn't require a dev envt on the computer to be effective - cross-vendor support - more sophistication for complex use cases (enterprise per…

I don't understand why tool calling isn't the primitive. A "skill" could have easily just been a tool that an agent can build an execute in its own compute space. I really don't see why we need two forms of RCP...

> A "skill" could have easily just been a tool

The problem skills solve is initial mapping of available information. A tool might hide what information it contains until used, this approach puts a table of contents for docs in the context, so the model is aware and can navigate to desired information as needed.

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#278

Post by Nevermark »

taormina wrote: post #268
hu3 wrote: post #264

Earlier quoted context omitted.

check your AGENTS.md or equivalent most of the times when LLM is misbehaving, it's my fault for leaving outdated instructions

Nah, it's definitely not that. I have it explicitly several times over in that file and it insists on commenting so much it's absurd.

Can you tell it to put the comments somewhere else? A special "Important Comments" folder? Ha! These things.

Re: Claude Skills are awesome, maybe a bigger deal than MCP

#280

Post by evanmoran »

Wowfunhappy wrote: post #252
7thpower wrote: post #232

Earlier quoted context omitted.

It is primarily a principal agent problem, with a hint of marshmallow test. If you are a developer who is not writing the documents for consumption by AI, you are primarily writing documents for someone who is not you; you do not know what this person will need or if they will ever even look at them. They may, of course, help you, but you may not understand that, have the time, or discipline. If you are writing them…

Your LLMs get rid of comments? Mine add them incessantly.

I know this is highly controversial, but I now leave the comments in. My theory is that the “probability space” the LLM is writing code in can’t help but write them, so if i leave them next LLM that reads the code will start in the same space. Maybe it’s too much, but currently I just want the code to be right and I’ve let go of the exact wording of comments/variables/types to move faster.
Post reply on HN