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

Claude Advanced Tool Use

anthropic.com

Post reply on HN
Print view View original

241–250 of 280 posts

Re: Claude Advanced Tool Use

#241

Post by esperent »

notpublic wrote: post #225
esperent wrote: post #161

Earlier quoted context omitted.

I haven't had much luck with skills being called appropriately. When I have a skill called "X doer", and then I write a prompt like "Open and do X", it almost never loads up the skill. I have to rewrite the prompt as "Open and do X using the X doer skill". Which is basically exactly as much effort as what I was doing previously of having prewritten sub-prompts/agents in files and loading up the file each time I want…

Try adding a session hook that triggers on startup|resume|clear|compact to remind Claude about your custom skills.

Is there a session start hook? I don't think so, unless it was added recently.

I've mostly been working on smaller projects so I never need to compact. And skills are definitely not working even on the initial prompt of a new session.

Re: Claude Advanced Tool Use

#242

Post by makestuff »

theknarf wrote: post #202

We should just build more CLI tools, that way the agentic AI can just run `yourtool --help` to learn how to use it. Instead of needing an MCP-server to access ex. Jira it should just call a cli tool `jira`. Better CLI tools for everything would help both AI and humans alike.

Yeah, I have a feeling we will instead start exposing some /help api that the AI will first call to see all possible operations and how to use them in some sort of minified documentation format.

Re: Claude Advanced Tool Use

#243

Post by throwaway19268 »

tomashubelbauer wrote: post #206
theknarf wrote: post #202

We should just build more CLI tools, that way the agentic AI can just run `yourtool --help` to learn how to use it. Instead of needing an MCP-server to access ex. Jira it should just call a cli tool `jira`. Better CLI tools for everything would help both AI and humans alike.

This would be awesome, but great CLIs would have already been valuable prior to the age of LLMs and yet most services didn't ship one. I think it is because services like Jira and others do not want to be too open. Ultimately, despite the current LLM/MCP craze, I think this won't change and MCP tools will start getting locked down and nerfed somehow, the same way APIs have in not so recent memory after there being a…

CLI tools for online services like Jira etc. basically amount to an open and documented API which the attitude towards these probably unlikely to be changing anytime soon as you mention.

Re: Claude Advanced Tool Use

#245

Post by losvedir »

jimbo808 wrote: post #154
losvedir wrote: post #80

I never really understood why you have to stuff all the tools in the context. Is there something wrong with having all your tools in, say, a markdown file, and having a subagent read it with a description of the problem at hand and returning just the tool needed at that moment? Is that what this tool search is?

Claude is pretty good at totally disregarding most of what’s in your CLAUDE.md, so I’m not optimistic. For example a project I work on gives it specific scripts to run when it runs automated tests, because the project is set up in a way that requires some special things to happen before tests will work correctly. I’ve never once seen it actually call those scripts on the first try. It always tries to run them using t…

That's kind of the opposite of what I mean. CLAUDE.md is (ostensibly) always loaded into the context window so it affects everything the model does.

I'm suggesting a POTENTIAL_TOOLS.md file that is not loaded into the context, but which Claude knows the existence of. That file would be an exhaustive list of all the tools you use, but which would be too many tokens to have perpetually in the context.

Finally, Claude would know - while it's planning - to invoke a sub-agent to read that file with a high level idea of what it wants to do, and let the sub-agent identify the subset of relevant tools and return those to the main agent. Since it was the sub-agent that evaluated the huge file, the main agent would only have the handful of relevant tools in its context.

Re: Claude Advanced Tool Use

#246

Post by linsomniac »

tomashubelbauer wrote: post #206
theknarf wrote: post #202

We should just build more CLI tools, that way the agentic AI can just run `yourtool --help` to learn how to use it. Instead of needing an MCP-server to access ex. Jira it should just call a cli tool `jira`. Better CLI tools for everything would help both AI and humans alike.

This would be awesome, but great CLIs would have already been valuable prior to the age of LLMs and yet most services didn't ship one. I think it is because services like Jira and others do not want to be too open. Ultimately, despite the current LLM/MCP craze, I think this won't change and MCP tools will start getting locked down and nerfed somehow, the same way APIs have in not so recent memory after there being a…

I've had good luck with having Claude write little CLI tools that interact with Jira: "cases" prints out a list of my in-progress cases (including immediately printing a cached list of cases, then querying Jira and showing any stragglers), "other_changes" shows me tickets in this release that are marked with "Other changes" label, "new_release" creates a new release in Jira, our deployment database, and a script to run the release, etc...

I could imagine a subagent that builds a tool on demand when it's needed.

Claude is really good at building small tools like these.

Re: Claude Advanced Tool Use

#247

Post by xnorswap »

pferde wrote: post #227
xnorswap wrote: post #226

Earlier quoted context omitted.

I agree with your conclusion that this stuff will get locked down again over time. I think there's also another major reason people don't like to ship desktop software, and that's the cost of support of dealing with outdated tools, it can be immense. Ticket as raised, "Why is my broken?" After several rounds of clarification, it's established they're using a 6-year old version that's hitting API endpoints that were f…

If that took "several rounds of clarification", then the support they're paying for is worthless. Getting version of the application should be among the first bits of information collected, possibly even required for opening the ticket.

You've never asked someone for a version and got back a version number for a completely different product?

Obviously it depends on your audience, and 3 rounds is exaggerating for the worst case, but in previous places I've worked I've seen customer support requests where the first question that needed to be asked wasn't, "What version are you using?", it's "Are you sure this is our product you're using?".

Actually getting version info out of that audience would have been at least an email explaining the exact steps, then possibly a follow up phone call to talk them through it and reassure them.

If your reference is JIRA tickets or you're selling software to software developers, then you're dealing with a heavily filtered stream. Ask your CS team for a look at the unfiltered incoming mail, it might be eye-opening if you've not done it before. You might be surprised just how much of their time is spent covering the absolute basics, often to people who have had the same support multiple times before.

Re: Claude Advanced Tool Use

#248

Post by chillfox »

theknarf wrote: post #202

We should just build more CLI tools, that way the agentic AI can just run `yourtool --help` to learn how to use it. Instead of needing an MCP-server to access ex. Jira it should just call a cli tool `jira`. Better CLI tools for everything would help both AI and humans alike.

That's pretty much how I have been using coding agents. I get them to build small cli tools with a --help option and place them in a `./tools` directory. Then I can tell an agent to use the tools to accomplish whatever task I need done.

Usually when I mention a `tool --help` for the first time in a prompt I will put it in backticks with the --help argument.

It works really well.

Re: Claude Advanced Tool Use

#249

Post by joshribakoff »

tomashubelbauer wrote: post #206
theknarf wrote: post #202

We should just build more CLI tools, that way the agentic AI can just run `yourtool --help` to learn how to use it. Instead of needing an MCP-server to access ex. Jira it should just call a cli tool `jira`. Better CLI tools for everything would help both AI and humans alike.

This would be awesome, but great CLIs would have already been valuable prior to the age of LLMs and yet most services didn't ship one. I think it is because services like Jira and others do not want to be too open. Ultimately, despite the current LLM/MCP craze, I think this won't change and MCP tools will start getting locked down and nerfed somehow, the same way APIs have in not so recent memory after there being a…

The good news is that an llm will be really helpful in scraping your content locating alternative service providers or even creating your own solution so you can migrate away
Post reply on HN