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

Agent Skills

addyosmani.com

Post reply on HN
Print view View original

181–190 of 239 posts

Re: Agent Skills

#181

Post by Chris2048 »

wg0 wrote: post #68

Snake oil. Good to read for sure. Seems all plausible too. But snake oil nevertheless. Here's why: The slot machine can drop any hard requirement that you specifically in your AGENTS.md, memory.md or your dozens of skill markdowns. Pretty much guaranteed. These harnesses approaches pretend as if LLMs are strict and perfect rule followers and the only problem is not being able to specify enough rules clearly enough. T…

> That leaves only one option not reliable but more reliable nevertheless: Human review and oversight.

Couldn't non-manual oversight also help e.g. sandboxes?

Re: Agent Skills

#182

Post by kergonath »

j45 wrote: post #177
kergonath wrote: post #162

Earlier quoted context omitted.

Indeed, and it is a complicated problem to solve. A GUI or CLI can hide footguns or make them less likely to be misused. But an AI agent is perfectly happy to use a wrecking ball to put a nail without any second thought or confirmation.

It’s a human articulation problem. When it receives a generic vague input it is free to interpret according to how its corpus fires like any human interaction. How to articulate better is like writing a sentence that will stand the test of model updates.

Even then. I don’t have an example off the top of my head but even perfectly clear sentences can lead the agent to strange places. Even between humans, miscommunication is easy, but then anyone sensible would ask for confirmation if their interpretation is weird. But the LLM very rarely questions the user.

I don’t think it’s fair to blame the user here. The tool must be operated by normal users.

Re: Agent Skills

#183

Post by fortyseven »

ai_fry_ur_brain wrote: post #39

Cant wait for everyone to realize they've wasted a year + messing with agents and experiencing a feeling of psuedo productivity.

I'm with the username like that, I'm sure we're going to get an even-handed, well thought out and reasoned discussion about all of this.

Re: Agent Skills

#184

Post by fortyseven »

dawnerd wrote: post #97
thatmf wrote: post #48

Why are people so excited to put themselves out of a job? Not that these or any "skills" will do that, but just- in principle. This is like alienation from labor at scale.

It's likely the people that were not good developers that suddenly got accelerated "to the top" that seem the most for it. All of the good devs I know have been a bit more cautious on the uptake.

Funny, I know quite a few extremely talented programmers who cautiously approached the topic, and found that, with proper use, they've found LLMs to be extremely useful. Just a matter of understanding where the boundaries are, and using them responsibly. It's not a magic genie, it augments their existing skill.

Re: Agent Skills

#185

Post by forlorn_mammoth »

sharperguy wrote: post #142
stellalo wrote: post #87

> A skill is a markdown file with frontmatter that gets injected into the agent’s context when the situation calls for it. When the LLM decides that the situation calls for it > It is a workflow: a sequence of steps the agent follows, with checkpoints that produce evidence, ending in a defined exit criterion. A sequence of steps the LLM can decide to follow

Skills are often invoked imperatively by the user. In cases where they are intended to be used directly by the LLM, it would be included somewhere else in the context. E.g: ``` After implementing the feature, read the testing skill for instructions on how to test. ```

how do you guarantee that the LLM follows an instruction given imperatively by the user? It probably will, but this is not guaranteed behavior. Likewise, _how_ it follows that instruction is non-deterministic.

it's turtles all the way down.

Re: Agent Skills

#187

Post by standardUser »

> It produces code, declares victory, and moves on.

Not when I'm in charge. It proposes changes based on my detailed instructions, I review the proposed changes, only then do I have it implement code, and then I review it again. I understand my AI agent would prefer a quicker way but for the meantime, I'm still the one in charge.

Re: Agent Skills

#188

Post by onlyrealcuzzo »

standardUser wrote: post #187

> It produces code, declares victory, and moves on. Not when I'm in charge. It proposes changes based on my detailed instructions, I review the proposed changes, only then do I have it implement code, and then I review it again. I understand my AI agent would prefer a quicker way but for the meantime, I'm still the one in charge.

I think you're saying the same thing OP said.

The point is, their default behavior is to ship crap fast.

You have a process to handle that.

So does OP.

Re: Agent Skills

#189

Post by onlyrealcuzzo »

thatmf wrote: post #48

Why are people so excited to put themselves out of a job? Not that these or any "skills" will do that, but just- in principle. This is like alienation from labor at scale.

We've been automating stuff for 60 years, and it only leads to more automation.

At the end of the day, the more automation, the more people you need making sure things work.

There's always going to be a minimal bottleneck for how much an engineer can oversee if they need to do zero implementation.

We're not as far from that point as people think.

Most languages most things are developed in are 10x more expensive than languages of yore.

Rust has a bad reputation for being hard, but it is actually quite expressive.

Less than 50% of what engineers do is code.

IBM was famous, in the early 2000s, for the average dev writing one line of code per day on average.

We're just going to move to a world where the average dev spends <10% of their time coding, but there's likely to be x times more work, so it mostly evens out.

Post reply on HN