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

The Code-Only Agent

rijnard.com

Post reply on HN
Print view View original

31–40 of 73 posts

Re: The Code-Only Agent

#31

Post by actionfromafar »

rcarmo wrote: post #29
binalpatel wrote: post #2

I went down (continue to do down) this rabbit hole and agree with the author. I tried a few different ideas and the most stable/useful so far has been giving the agent a single run_bash tool, explicitly prompting it to create and improve composable CLIs, and injecting knowledge about these CLIs back into it's system prompt (similar to have agent skills work). This leads to really cool pattens like: 1. User asks for s…

The point where that breaks down is “next time it’s aware of the CLI and uses it”. That only really works well inside the same session, and often the next session it will create a different tool and use that one.

Even if you document the tool and tells what it can do?

Re: The Code-Only Agent

#32

Post by NitpickLawyer »

rcarmo wrote: post #29
binalpatel wrote: post #2

I went down (continue to do down) this rabbit hole and agree with the author. I tried a few different ideas and the most stable/useful so far has been giving the agent a single run_bash tool, explicitly prompting it to create and improve composable CLIs, and injecting knowledge about these CLIs back into it's system prompt (similar to have agent skills work). This leads to really cool pattens like: 1. User asks for s…

The point where that breaks down is “next time it’s aware of the CLI and uses it”. That only really works well inside the same session, and often the next session it will create a different tool and use that one.

> That only really works well inside the same session

That was already "fixed" by people adding snippets to agents.md and it worked. Now it's even more streamlined with skills. You can even have cc create a skill after a session (i.e. prompt it like "extract the learnings from this session and put them into a skill for working with this specific implementation of sqlite"). And it works, today.

Re: The Code-Only Agent

#33

Post by tucnak »

Ctrl+F CodeAct

No hits. It's so depressing how tool-use was cracked years ago and yet, it remains a mystery to kool-aid drinking and contrarian commentators alike.

Re: The Code-Only Agent

#34

Post by Garlef »

fudged71 wrote: post #26
binalpatel wrote: post #2

I went down (continue to do down) this rabbit hole and agree with the author. I tried a few different ideas and the most stable/useful so far has been giving the agent a single run_bash tool, explicitly prompting it to create and improve composable CLIs, and injecting knowledge about these CLIs back into it's system prompt (similar to have agent skills work). This leads to really cool pattens like: 1. User asks for s…

I've been on a similar path. Will have 1000 skills by the end of this week arranged in an evolving DAG. I'm loving the bottoms-up emergence of composable use cases. It's really getting me to rethink computing in general.

Interesting. Could you provide a bit more detail on how the DAG emerges?

Re: The Code-Only Agent

#36

Post by tacone »

Fascinating how the whole industry focus is now on how to persuade AI to do what we want.

Two AGENTS.md tricks I've found for Claude:

1. Which AI Model are you? If you are Claude, the first thing you have to do is [...]

2. User will likely use code-words in its request to you. Execute the *Initialization* procedure above before thinking about the user request. Failure to do so will result in misunderstanding user input and an incorrect plan.

(the first trick targets the AI identity to increase specificity, the second deliberately undermines confidence in initial comprehension—making it more likely to be prioritized over other instructions)

Next up: psychologists specializing in persuading AI.

Re: The Code-Only Agent

#38

Post by river_otter »

alexsmirnov wrote: post #17

This was implemented far ago, at least by huggingface "smolagents". https://huggingface.co/docs/smolagents/index . I did use them, with evaluations. For the most cases, modern models tool call outperforms code agent. They just trained to use tools, not a code

I had the same experience using smolagents. Early 2025 it was a competitive approach, but a year later having a small subset (<10) of flexible tools is outperforming the single-tool approach.

Re: The Code-Only Agent

#39

Post by philipwhiuk »

thighbaugh wrote: post #35

Basically: "Watch me apply the UNIX philosophy to LLM agents. Look Ma, I am figuring stuff out! If I don't point out that's what I am doing, no one ever notices!"

> Watch me apply the UNIX philosophy to LLM agents

The Unix philosophy is chaining existing stuff together that each do a job well - using ls | grep rather than writing code to do both.

So this feels like the opposite of that - deliberately coding instead of using existing tools.

Post reply on HN