Rendered at 08:00:53 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
MikhailTal 3 hours ago [-]
Very fascinating, super interesting engineering. Although i do find it very funny how they just bypass a massive vulnerability, basically zero data isolation (even between good actors, let alone bad ones) with 3 sentences. Only in the llm space you can slap a massive limitation like this in the middle of the article and continue like nothing happened
> Whatever anyone tells Audel becomes part of the single experience that every other conversation draws on. In practice, Audel is bad at keeping secrets. Ask it what it’s been working on with someone else and it will often just tell you, even though we’ve asked it not to. We also haven’t studied what happens when two people give conflicting instructions. For now, we assume anything you tell Audel is shared with everyone on the team.
vedtam 55 minutes ago [-]
"turn -> FINAL -> schedule wake-up", this is where my excitement has faded unfortunately. Many of us are probably wondering about the same idea: bridging the gap between a reactive agent and my daily workflow or existence. But, this still feels too close to how Claude (or any other agent) runs as a process in the background (always ON), where you can use a custom channel to feed the dialog with external signals like chat, CI/CD events, whatsapp, etc.
Humans aren't scheduling a wake-up to the next thought. Ideally, a sub second agentic loop with no FINAL / wake-up, always "spinning" would get closer. I'm conscious about the waste of resources this would drag with it (because of current architectures), but exciting still.
PS. I love the take on using bash instead of Python (one less abstraction layer!) and using UNIX fundamentals as stepping stone when composing tools as agents are naturally drawn to using it on a box anyways.
pm25coder 41 minutes ago [-]
[flagged]
weinzierl 41 minutes ago [-]
The language composition is interesting. The source is half Shell, a quarter Python, almost a fifth Typescript. Among the rest is 2.3% Rust and 1.3% Swift.
If you wonder what the Rust is for: It is the Ratatui TUI.
yewenjie 4 hours ago [-]
Are there any objective metrics/ benchmarks that people test harnesses by?
There are just so many now that it's hard to personally test them all or just trust the vibes.
andyk 2 hours ago [-]
andy here (headlong post author).
terminal bench 3 is pretty popular for comparing different harnesses using the same underlying model (it's another laude project actually). artificial analysis has an index. you can look at the model cards of popular model releases- they tend to have the most popular current benchmarks on them.
w/ headlong we decided to announce it before we've benchmarked it. we mostly wanted to informally share our experiences w/ it in this initial post. we plan to do some benchmarking coming up here soon tho
weinzierl 48 minutes ago [-]
The page has a nice little easter egg if you click "Dr. K" in the footer.
airocker 3 hours ago [-]
Sub Question : IS there a real successful agent product today that uses a library for harness(like langgraph etc)? Building our own worked for us. Works with our components(postgres, events ...) and scales naturally with our system.
gexla 2 hours ago [-]
I don't know about real successful. Since you mentioned Langchain, you could look at https://www.langchain.com/dcode which is a CLI harness build off Langchain deep agents.
3 hours ago [-]
imagetic 55 minutes ago [-]
that imaginary line. this crossed it.
0xbadcafebee 2 hours ago [-]
> Audel designed experiments to spawn recursive shellm sub-runs to work on subproblems. Most of the experiments failed, because shellm has a safety watchdog that kills any command that stays silent for 30 seconds. Audel fought the watchdog for about 40 minutes and mostly stopped using shellm sub-runs. Results from recursive sub-runs of shellm merged back into Audel’s mind 64 times in its first two days and 12 times in the twelve days since. We’ve since revamped the watchdog, and we’ll see if we can convince Audel to give recursion another shot.
This is why "I made it think in a loop" doesn't result in significant improvement in LLM performance. It's not learning. You need RLAIF, STAR, IDPO, etc to retrain the model to learn from its mistakes. And you need a human to review it so it's not compounding mistakes. It's expensive and time-consuming. Doing it wrong leads to bad outcomes. But not doing it leads to no significant improvement.
JacobAsmuth 3 hours ago [-]
The Googlers must be vague posting about something internal.
jnwatson 3 hours ago [-]
It buries the lede. Prime Agent sounds like a very cool project.
ma2kx 2 hours ago [-]
I'm just exhausted. So I've today now learned about four new harness:
Of course the don't have exactly the same scopes but they are in general all about persistent memory and / or continous agent loops. Like I miss those times where only once a week a new js framework was promoted.
simianwords 56 minutes ago [-]
Why compress by recency rather than something else?
russellbeattie 3 hours ago [-]
> "Headlong is a complete agent harness with a core of less than 10K lines of Bash..."
Wow. So, be nice or I'll replace you with a very large shell script?
> Whatever anyone tells Audel becomes part of the single experience that every other conversation draws on. In practice, Audel is bad at keeping secrets. Ask it what it’s been working on with someone else and it will often just tell you, even though we’ve asked it not to. We also haven’t studied what happens when two people give conflicting instructions. For now, we assume anything you tell Audel is shared with everyone on the team.
Humans aren't scheduling a wake-up to the next thought. Ideally, a sub second agentic loop with no FINAL / wake-up, always "spinning" would get closer. I'm conscious about the waste of resources this would drag with it (because of current architectures), but exciting still.
PS. I love the take on using bash instead of Python (one less abstraction layer!) and using UNIX fundamentals as stepping stone when composing tools as agents are naturally drawn to using it on a box anyways.
If you wonder what the Rust is for: It is the Ratatui TUI.
There are just so many now that it's hard to personally test them all or just trust the vibes.
This is why "I made it think in a loop" doesn't result in significant improvement in LLM performance. It's not learning. You need RLAIF, STAR, IDPO, etc to retrain the model to learn from its mistakes. And you need a human to review it so it's not compounding mistakes. It's expensive and time-consuming. Doing it wrong leads to bad outcomes. But not doing it leads to no significant improvement.
https://github.com/exoharness/exo/
https://github.com/laude-institute/headlong
https://github.com/microsoft/agent-lightning
and now https://github.com/PrimeIntellect-ai/prime-agent
Of course the don't have exactly the same scopes but they are in general all about persistent memory and / or continous agent loops. Like I miss those times where only once a week a new js framework was promoted.
Wow. So, be nice or I'll replace you with a very large shell script?