frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Escape IntelliJ: Scala and Kotlin LSPs on Emacs Eglot

https://jointhefreeworld.org/blog/articles/emacs/emacs-eglot-scala-kotlin/index.html
87•jjba23•2d ago•55 comments

Terence Tao's ChatGPT conversation about the Jacobian Conjecture counterexample

https://chatgpt.com/share/6a5fdc7a-d6f8-83e8-bbea-8deb42cfed56
932•gmays•18h ago•538 comments

Cruller: Bun's Zig Runtime, Continued on Zig 0.16

https://ziggit.dev/t/cruller-buns-zig-runtime-continued-on-zig-0-16/16734
74•Erenay09•6h ago•37 comments

Quality non-fiction books are the antithesis of AI slop

https://resobscura.substack.com/p/quality-non-fiction-books-are-the
397•benbreen•21h ago•155 comments

EU fines Google €890M for competition breaches over search and apps

https://www.theguardian.com/technology/2026/jul/23/eu-fines-google-for-competition-breaches-over-...
61•Stevvo•1h ago•57 comments

Code mode yields a 99.2% cost reduction in our systems

https://www.agent-swarm.dev/blog/code-mode-token-savings
54•tarasyarema•1h ago•42 comments

The Unity CLI: manage Unity from your terminal

https://unity.com/blog/meet-the-unity-cli
15•nateb2022•1d ago•2 comments

GigaToken: ~1000x faster Language model tokenization

https://github.com/marcelroed/gigatoken/
533•syrusakbary•18h ago•113 comments

Show HN: Bento - An entire PowerPoint in one HTML file (edit+view+data+collab)

https://bento.page/slides/
878•starfallg•20h ago•197 comments

Everyone should know SIMD

https://mitchellh.com/writing/everyone-should-know-simd
473•WadeGrimridge•18h ago•174 comments

Worse on Purpose – How Corporate Greed Killed Product Quality – Worse on Purpose

https://www.worseonpurpose.com/
32•bilsbie•55m ago•11 comments

New Framework Desktop Option with AMD Ryzen AI Max+ Pro 495 and 192GB Memory

https://frame.work/desktop?tab=192gb-coming-soon
19•PhilippGille•51m ago•4 comments

Are AI labs pelicanmaxxing?

https://dylancastillo.co/posts/pelicanmaxxing.html
576•dcastm•18h ago•222 comments

ANSI escape injection in MCP servers: Hidden from humans, visible to AI

https://brightsec.com/research/detecting-ansi-escape-sequence-injection-in-mcp-servers-with-dast/
27•xgpyc2qp•2d ago•11 comments

So Reddit has decided that plain HTML is unsafe

https://www.cole-k.com/2026/07/21/reddit/
513•montroser•23h ago•514 comments

git's –end-of-options Flag

https://nesbitt.io/2026/07/21/end-of-options.html
163•Erenay09•1d ago•97 comments

Protecting our FLOSS commons from LLMs

https://blog.codeberg.org/protecting-our-floss-commons-from-llms.html
92•acmnrs•10h ago•33 comments

Amiga 1000: Ten years ahead of its time

https://dfarq.homeip.net/amiga-1000-ten-years-ahead-of-its-time/
104•giuliomagnifico•6h ago•89 comments

Making

https://beej.us/blog/data/ai-making/
382•erikschoster•20h ago•150 comments

The startup's Postgres survival guide

https://hatchet.run/blog/postgres-survival-guide
431•abelanger•23h ago•197 comments

Show HN: Cactus Hybrid: We taught Gemma 4 to know when it's wrong

https://github.com/cactus-compute/cactus-hybrid
152•HenryNdubuaku•18h ago•35 comments

Making ASCII Art in Vim

https://alexyang.dev/vim-ascii-art/
80•evakhoury•2d ago•8 comments

Medici family mystery may be solved after more than 400 years

https://www.cnn.com/2026/07/15/science/medici-family-mystery-dna-malaria
131•effects•14h ago•38 comments

John C. Dvorak has died

https://twitter.com/na_announce/status/2079952538040672302
790•coleca•16h ago•265 comments

Why malloc always does more than I asked for?

https://ssenthilnathan3.github.io/blog/malloc/
44•nathaah3•3d ago•36 comments

Businesses with ugly AI menu redesigns

https://blog.fiddery.com/businesses-with-ugly-ai-menu-redesigns/
321•speckx•23h ago•232 comments

Malleable Computing, Emacs, and You

http://yummymelon.com/devnull/malleable-computing-emacs-and-you.html
118•kickingvegas•14h ago•33 comments

Frequently Asked Questions on Expertise

https://jtpeterson.substack.com/p/faq-on-expertise
12•surprisetalk•2d ago•0 comments

Nobody knows what a used GPU cluster is worth

https://ciphertalk.substack.com/p/nobody-knows-what-a-used-gpu-cluster
245•rbanffy•1w ago•220 comments

Fairphone 6 wide camera experimental Linux support

https://nondescriptpointer.com/articles/fairphone-6-wide-camera-linux/
134•helonaut•15h ago•42 comments
Open in hackernews

Code mode yields a 99.2% cost reduction in our systems

https://www.agent-swarm.dev/blog/code-mode-token-savings
50•tarasyarema•1h ago

Comments

tarasyarema•1h ago
Hi there!

We've been building https://github.com/desplega-ai/agent-swarm in the open for a while, and one of the things we wanted to do since the start is find how powerful it can be to push for a code mode like env for the swarm.

We tried it and we managed to reduce up +90% of token costs by using swarm scripts (the code mode variant we implemented) for some of our schedules.

This is game changer, as we are able to

1. Run those schedules faster 2. Cheaper 3. More reliably

Been thinking a lot now on how we should prompt and change the default templates to force the agents to build on top of this. If it can be re-used or it's a recurring thing -> scripts.

Thoughts? Have you seen this type of improvements in your setups?

Cheers,

mpalmer•30m ago

    We didn't add this for the post. It's just what the swarm already does, and it's the same machinery behind Script Workflows.

    What we hadn't done is measure it against Anthropic's own yardstick, with our own production data.
You say here that "since the start" you wanted to do a "code mode like env".

But the LLM that wrote your blog post says your system was already doing this and you just hadn't measured it...? How is it that you managed a reduction in token costs by changing nothing?

And can you provide a human-written explanation of the experimental methodology that gets you this miraculous, literally unbelievable 99.2% reduction in tokens?

tarasyarema•21m ago
Hi, the blog post was written assisted with our own swarm, based on the experiment.

Indeed we had the concept of workflows from the start (like n8n), but the point of the scripts is type safe executions that the agents can write, with access to APIs, MCPs, etc.

E.g. this PR from ~10 days ago https://github.com/desplega-ai/agent-swarm/pull/934 was going in this direction, offering a way to defined type safe connections to able to use more the scripts instead of adding MCPs or tools.

e.g. offering a way to do `ctx.api.gmail.searchEmail` coming from an OpenAPI spec.

The whole point of this blog is the "dumb" realisation of the following:

If you have an agent (say Claude Tag, or some other fancy Slack bot) that you keep asking the same things (e.g. check DataDog error and correlate PRs on a timeframe) and the agent does that process agentically, you are wasting tokens.

Now, if you give a way for the agents to write these scripts (code mode) the next time a task like that happens then essentially you just run that and you get easily 90% cost reduction in THAT process. Not globally, obviously.

moberemk•15m ago
No offence but, “if you find yourself doing the same thing over and over again write a script for it” isn’t exactly novel thinking in the software field…this was the same process when it was a human doing it, we just measure that in time instead of tokens
tarasyarema•10m ago
yes, 100%

but it's not that simple. the whole point of this is that if you have an AI system that is in charge of this, what you need is a way for it to build those deterministic software based parts, while maintaining the AI routing and reasoning.

with time, our internal swarm has been building those "internal" software that is tedious and boring to build, where it makes sense.

e.g. we also have a "pages" thing (like claude arifacts) that it uses to join scripts, html and other components to build dashboards for us

lyall•19m ago
There might be something valuable in this article, but if the you couldn't be bothered to write it yourself, why should we be bothered to read it?
tarasyarema•12m ago
This is based on a multi-day convo I had with our swarm. I might have not wrote it myself, but it's based on human interaction and thought process.

Any recommendations on how I could make it better next time, so you do not feel like wasting your time (which I get and feel, dont get me wrong!)

somnium_sn•37m ago
I am still a bit surprised that this is not more widely used. Normal inference + tool calls is effectively composition via serialized natural language, whereas Code Mode / Programmatic Tool Calling is using purpose built languages (aka programming languages) for composition and concurrency. It somewhat feels obvious to me that the latter is way more token and latency efficient.
asdfsa32•27m ago
In the world of AI hype, everything is to be learned again and we are going to see many many breakthroughs like this. You can see this shift already in the tone of conversation in HN for example.

The basics of computer engineering is now novel approaches. We are discovering why programming languages exist, starting with a worst version of Cobal running on a very expensive and probabilistic VM.

We will shoehorn every square, rectangular, and all forms of peg through the AI hole.

jorisw•16m ago
TFA suggests that it already is:

> The one-line takeaway

> This already ships by default in every claude/codex/opencode session's system prompt

shireboy•30m ago
A maxim I’ve arrived at working w llm every day is “put deterministic things in code, non deterministic in llm”. I do wish the harnesses would be more helpful in this regard. For example I’ve seen tons of scheduled jobs that people wrote in Claude/copilot/etc that could easily have been scripts. They aren’t scripts because the author doesn’t know how to script and they stop at the point the llm does what they want. It isn’t hard to then tell the llm “make a script to do all of the tool calls this prompt needs and then pass the result back to the llm for this non-deterministic part”.
j45•6m ago
I’d expect the llm to write the script and run it.

The tooling can remain generic and token spending focused.

fg137•28m ago
The article is so painful to read.
stingraycharles•24m ago
Yes and it’s also something that was already known and recommended. It’s not “code mode” or something revolutionary, it’s just instructing your LLM to feel free to write scripts to aid it in completing the task.

It’s very useful and I use it all the time, and in the end you can distill it into a reusable toolbox, somewhat similar to a collection of markdown files with project knowledge.

tarasyarema•7m ago
still most harnesses and AI systems do not use it, adopt it, or reference it enough...
elzbardico•23m ago
Because it was probably written by an LLM.

Use a LLM to summarize it.

Hamuko•18m ago
>Use a LLM to summarize it.

GIGO?

VulgarExigency•6m ago
Relevant Slavoj Zizek talk: https://www.youtube.com/watch?v=qj4cxK-nOTw
vessenes•26m ago
This quarter’s clankerisms: “Honest” and “vibes”.

The tech here is cool; good reminder to push for more scripts in general from the agentic tools.

tarasyarema•9m ago
thanks!
AmazingTurtle•26m ago
codex CLI is already leveraging this internally. instead of performing raw MCP calls its internally using a js REPL to do that. elicitation is then propagated for approvals independently / it's interpreted as regular tool calls for operators etc.

so if i were to tell an agent to move the contents of a confluence doc into a file, it would do so without even reading the confluence page - in theory.

tarasyarema•13m ago
Yes, that's the right direction codex is taking 100%

Now our idea is to offer this to be harness agnostic (we support codex, claude code, pi, opencode, even devin). And you (as the one deploying AI systems) should be in control of the AI, the execution and the data. That's our goal

grzracz•24m ago
I am once again asking for people to ask Claude to reformat their article text to ASD-STE100 so that I don't have to read slop
tarasyarema•10m ago
will do next time, thanks!
firasd•21m ago
The problem is that when the LLM writes ad-hoc code how can you trust it?

For example in my Liveclip MCP server I'm working on (not released yet) I have table manipulation type tools

So for example a couple days ago Claude made a combined ranking score for my Youtube Shorts analytics -- it did all these tool calls in the same turn

First it normalized the % values into numbers

{ "destination_col": "G", "key": "cinemasleepstories_temp_2026071702", "pattern": "%", "replacement": "", "source": { "col": "E", "row_start": 1, "row_end": 6 } }

Then it made the composite score

{ "dest_col_start": "H", "dest_row_end": 6, "dest_row_start": 1, "expr": "round((likes+1)(followers+1)stayed, 1)", "source_key": "cinemasleepstories_temp_2026071702", "sources": [ { "col": "C", "name": "likes", "row_offset": 0 }, { "col": "D", "name": "followers", "row_offset": 0 }, { "col": "G", "name": "stayed", "row_offset": 0 } ] }

And then it set the new headers

{ "headers": { "G": "Stayed (numeric)", "H": "Composite Score" }, "key": "cinemasleepstories_temp_2026071702" }

If Claude just said "let me write some pandas against this CSV" the workflow would be a lot more iffy and generally uncomfortable/ephemeral

asdfsa32•18m ago
A python script is now more iffy than a LLM going through JSON?
firasd•14m ago
Yeah, of course. The MCP server tools do use Python in the backend but just look at the affordances it provided in the args -- named columns, a fixed match/replace API, a way to specify the range of rows. The LLM is not gonna create a little library with these affordances in an ephemeral REPL it's just gonna YOLO some pandas code that you hope did what you wanted and addressed the right columns, had the right types, etc
tarasyarema•
ramon156•18m ago
LLM-assisted websites yield a 99.2% trust reduction in me
tarasyarema•15m ago
I get it, but you got it right. It was assisted, not just AI slop. Also nice that it's not 100% :D
vanyaland•12m ago
you lose the per-call approvals though. one script hitting ten tools is harder to gate than ten separate calls
tarasyarema•8m ago
yes, that is true.

in fact we are thinking a lot about this, not on a toll based, but rather on a role based.

giving access to specific tools to specific agents, how that works and evolves, and how that links w the humans using it

Planktonne•9m ago
I could be misreading this, but looking at the numbers, it seems that--despite the claim "we didn't estimate this"--basically everything in this is an estimate. It's a little hard to tell though, because the generated prose seems to occasionally contradict or repeat itself.
Gabrys1•8m ago
I read it like that: we haven't measured this YET until now, and now we did
tarasyarema•4m ago
thanks for the feedback!

it's really hard to measure some times. for example, once of the things we have seen is that the specific `workflow-triage` schedule we have have been running much faster and cheaper since we moved some parts of it to use scripts. From our analytics we saw times from ~5min before to ~1 or 2 min now. And in terms of costs we saw +50% reduction.

Why not 90%? because even if it uses the script, that's the "repeatable" part. we still have a reasoning around the output of it, and then deciding how to escalate, which removes the 40% reduction. Still 50% is nice for something we run daily.

16m ago
we have a way to see created scripts in a dashboard, and in fact they are not ad-hoc only, i.e. it can create "saved scripts" that it can re-use, iterate and so on.

Also not exec happens in a sandboxed env, which means it might not have access to the fs by default (it has to agent-fs, which is a layer we built on top of s3 for agents)

firasd•6m ago
Yeah I guess the dichotomy between 'tools' and 'code mode' isn't really as clear especially if the tools themselves are intelligent and truncate long results and have start_range, end_range type arguments. So the gist is about 'what part of the output needs to reach the LLM' and a 'code mode' approach can help orchestrate that just like good tools can