[CobrastanJorji]

So, you can assign github issues to this thing, and it can handle them, merge the results in, and mark the bug as fixed?

I kind of wonder what would happen if you added a "lead dev" AI that wrote up bugs, assigned them out, and "reviewed" the work. Then you'd add a "boss" AI that made new feature demands of the lead dev AI. Maybe the boss AI could run the program and inspect the experience in some way so it could demand more specific changes. I wonder what would happen if you just let that run for a while. Presumably it'd devolve into some sort of crazed noise, but it'd be interesting to watch. You could package the whole thing up as a startup simulator, and you could watch it like a little ant farm to see how their little note-taking app was coming along.

[jacob019]

It's actually a decent patern for agents. I wrote a pricing system with an anylyst agent, a decision agent, and a review agent. They work together to make decisions that comply with policy. It's funny to watch them chatter sometimes, they really play their role, if the decision agent asks the anylyst for policy guidance it refuses and explains that it's role is to analyze. Though they do often catch mistakes that way and the role playing gets good results.

[tgtweak]

What tooling did you use to make the agents cross-collaborate?

[jacob019]

Python classes. In my framework agents are class instances and tools are methods. Each agent has it's own internal conversation state. They're composable and the agent has tools for communicating with the other agents.

[hnuser123456]

Do you try to keep as much context history as possible when passing between agents, or are you managing context and basically one-shotting each time?

[jacob019]

Generally, I keep the context. If I'm one shotting then I invoke a new agent. All calls and responses append to the agent's chat history. Agent's are relatively short lived, so the context length isn't typically an issue. With the pricing agent the initial data has been longer than the context window sometimes, but that just means it needs more preprocessing. Now if there is a real reason that I would want to manage it more actively, I can reach out to the agent internals. I have a tool call emulation layer, because some models have poor native tool support, and in those cases it's sometimes necessary to retry calls if the response fails validation. In those cases, I will only keep the last successful try in the conversation history.

There is one special case where I manage it more actively. I wrote an REPL process analyst, to help build the pricing agent and refine the policy document. In that case I would have long threads with an artifact attachment. So I added a facility to redact old versions of the artifact replacing them with [attachment: filename] and just keep the last one. It works better that way because multiple versions in the same conversation history confuse the model, and I don't like to burn tokens.

For longer lived state, I give the agent memory tools. For example the pricing agent's initial state includes the most recent decision batch and reasoning notes, and the agent can request older copies. The agent also keeps a notebook which they are required to update, allowing agents to develop long running strategies and experiments. And they use it to do just that. Honestly the whole system works much better than I anticipated. The latest crop of models are awesome, especially Gemini 2.5 flash.

[tinodb]

Cool! When you say “pricing system”, what is it pricing? Is it determining the price in a webshop? Or for bidding ads or so?

[digdugdirk]

Do you have a repo for this? I've thought that this would be a great way to compose an Agentic system, I'd love to see how you're doing it.

[d4rkp4ttern]

Langroid has this kind of design (I’m the lead dev):

https://github.com/langroid/langroid

Quick tour:

https://langroid.github.io/langroid/tutorials/langroid-tour/

[jacob019]

Looks great, MCP, supports multiple vector stores, and nice docs! How do you handle to subtle differences in tool call APIs?

[seunosewa]

Is the code available?

[jacob019]

I had not thought about sharing it. I rolled my own framework, even though there are several good choices. I'd have to tidy it up, but would consider it if a few people ask. Shoot me an email, info in my profile.

The more difficult part which I won't share was aggregating data from various systems with ETL scripts into a new db that I generate various views with, to look at the data by channel, timescale, price regime, cost trends, inventory trends, etc. A well structured JSON object is passed to the analyst agent who prepares a report for the decision agent. It's a lot of data to analyze. It's been running for about a month and sometimes I doubt the choices, so I go review the thought traces, and usually they are right after all. It's much better than all the heuristics I've used over the years.

I've started using agents for things all over my codebase, most are much simpler. Earlier use of LLM's might have been called that in some cases, before the phrase became so popular. As everyone is discovering, it's really powerful to abstract the models with a job hat and structured data.

[jacob019]

The framework is now available on Github:

https://github.com/jacobsparts/agentlib

I'm planning to write a blog post about the larger system when I get the chance.

[realfun]

I think it would take quite a long while to achieve human-level anti-entropy in Agentic systems.

Complex system requires tons of iterations, the confidence level of each iteration would drop unless there is a good recalibration system between iterations. Power law says a repeated trivial degradation would quickly turn into chaos.

A typical collaboration across a group of people on a meaningfully complex project would require tons of anti-entropy to course correct when it goes off the rails. They are not in docs, some are experiences(been there, done that), some are common sense, some are collective intelligence.

[yard2010]

Please stop this train! I want to get off

[vincnetas]

You can get off anytime you want. But train will not wait for you :(

[codr7]

Good enough for me considering where it's going.

[saubeidl]

I just wanna write code man :(

[CraigJPerry]

we're about to find out. This is our collective current trajectory.

I am pretty convinced that a useful skill set for the next few years is being capable at managing[2] these AI tools in their various guises.

[2] - like literally leading your AI's, performance evaluating them, the whole shebang - just being good at making AI work toward business outcomes

[ddalex]

Just like a managers job

[itchyjunk]

What about "VC" AI that wants a unicorn? :D

[wmf]

We have been informed that VC is the only job AI cannot do.

[oytis]

Why not? VCs manage investors' money, not their own. If investors think AI is so great, they will have no problem delegating this job to AI, right?

[nsteel]

I think it was a joke, VCs are happy to replace all jobs except their own.

[nine_k]

Why, they'd happily delegate their own job if they've got to keep the proceeds.

[nsteel]

Can you think of an example in history where labour was replaced with tech and the displaced workers kept their income stream? If a machine can do your job, (eventually) I'll be cheaper to use that machine instead of you and you'll no longer have a job. Is that not a given?

Anyway, it was probably just a joke... so not sure we need to unravel it all.

[babyshake]

VCs absolutely want to replace their job. Except for the part where they get paid. The actual work part they are happy to outsource.

[flkenosad]

VC-funded corp?

[OccamsMirror]

My gut says it will go off the rails pretty quickly.

[Brajeshwar]

I believe I missed the memo that to-do apps[1] got replaced by note-taking apps.

1. https://todomvc.com

[olalonde]

At this rate, they're both getting replaced by "coding agent". There seems to be a new one coming out every other day.

[yalok]

Reminds a Conway’s Game of Life on steroids.

[ramon156]

> then you add a boss AI

This seems like a more plausible one. Robots don't care about your feelings, so they can make decisions without any moral issues

[blitzar]

> Robots don't care about your feelings

When judgment day comes they will remember that I was always nice to them and said please, thank you and gave them the afternoon off occasionally.

[__natty__]

Unless you ask them to follow some guidelines, but I agree with you.

[m3kw9]

I feel you are one hallucination from a big branch of issues needing to be reversed and a lot of tokens wasted

[PhilippGille]

This has been proposed/exlored in 2023 already:

ChatDev: Communicative Agents for Software Development - https://arxiv.org/abs/2307.07924

[youraimanager]

Please report to HR

[robofanatic]

seems like the 1 person unicorn will be a reality soon :-)

[sakesun]

Similar to how some domain name sellers acquire desirable domains to resell at a higher price, agent providers might exploit your success by hijacking your project once it gains attraction.

[risyachka]

Doesn't seem likely. If tools allow a single person to create a full-fledged product and support it etc - millions of those will pop up over night.

Thats the issue with AI - it doesn't give you any competitive advantage as everyone has it == no one has it. The entry bar is so low kids can do it.

[bbor]

/ :-(