frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Running AI agents across environments needs a proper solution

https://github.com/liquidos-ai/Odyssey
7•human_hack3r•1h ago
Hi HN folks,

I have been building AI agents for quite some time now. The shift has gone from LLM + Tools → LLM Workflows → Agent + Tools + Memory, and now we are finally seeing true agency emerge: agents as systems composed of tools, command-line access, fine-grained system capabilities, and memory.

This way of building agents is powerful, and I believe it is here to stay. But the real question is: are the systems powering these agents ready for that future?

I do not think so.

Using Docker for a single agent is not going to scale well, because agents need to be lightweight and fast. LLMs already add significant latency, so adding heavy runtime overhead on top only makes things worse. Existing solutions start to fall apart here.

Agents built in Python also tend to have a large memory footprint, which becomes a serious problem when you want to scale to thousands of agents.

And open-source for agents is still not where it should be. Right now, I cannot easily reuse agents built by domain experts the same way I reuse open-source software.

These issues bothered me, and I realized that if agents are ever going to be democratized, they need to be open and easy to use. Just like Docker solved system dependencies, we need something similar for agents.

That is why I started building an agent framework in Rust. It is modular and follows the principle of true agency: an agent is an entity with tools, memory, and an executor. In AutoAgents, users can independently create and modify tools, executors, and memory.

With AutoAgents, I saw that powerful agents could be built without compromising on performance or memory the way many other frameworks do.

But the other problems still remained: re-sharing agents, sandboxing, and scaling to thousands of agents.

So I created Odyssey — a bundle-first agent runtime written in Rust on top of AutoAgents, the Rust agent framework. It lets you define an agent once, package it as a portable artifact, and run it through the same execution model across local development, embedded SDK usage, shared runtime servers, and terminal workflows.

Both AutoAgents and Odyssey are fully open source and built in Rust, and I am planning to build an Odyssey Agent Hub soon, with additional features like WASM tools, custom memory layers, and more.

My vision is to democratize agents so they are available to everyone — securely and performantly. Being open is not enough; agents also need to be secure.

The project is still in alpha, but it is in a working state.

AutoAgents Repo -> https://github.com/liquidos-ai/AutoAgents

Odyssey Repo -> https://github.com/liquidos-ai/Odyssey

I would really appreciate feedback — especially from anyone who has dealt with similar problems. Your feedback help me shape the product.

Thanks for your time in advance!

Comments

harshitkuppa26•1h ago
Great work!!
rvz•1h ago
What were the insider tricks that you learned [0] to bot this post? You know using AI to do that here is against the guidelines in [1]?

Also, why is there one new (green) account created 10 - 15 mins ago close to the time of this post for the purpose of boosting it?

[0] https://news.ycombinator.com/item?id=47229114

[1] https://news.ycombinator.com/newsguidelines.html

human_hack3r•1h ago
[0] - No insider tricks, I posted couple of times every time it was dead post. This time I waited for couple days and then posting it again.

Not sure on the green account.

prernagoyal976•58m ago
Looks interesting!!
iatenice•13m ago
Superb

Create Context Graph: AI agents with graph based memory, scaffolded in seconds

https://github.com/neo4j-labs/create-context-graph
1•johnymontana•36s ago•0 comments

Sudo hold me – Three unsupervised LLMs on a Unix Socket

https://linuxtoaster.com/blog/emergent.html
1•dirk94018•58s ago•0 comments

D for learning and real world D: D Language Yale Univ Colloquium 2025 [video]

https://www.youtube.com/watch?v=k5r8d5KCOBE
1•teleforce•1m ago•0 comments

Show HN: ClawX – OpenClaw Instance in 30s

1•Yogender78•1m ago•0 comments

AMD GPU-Initiated I/O (GPU <-> NVMe P2P DMA)

https://thegeeko.me/blog/nvme-amdgpu-p2pdma/
1•thegeeko•1m ago•1 comments

Firefox 149.0 Release Notes

https://www.firefox.com/en-US/firefox/149.0/releasenotes/
1•N19PEDL2•2m ago•0 comments

Usecomputer – Zig CLI to screenshot, click and type in any coding agent

https://github.com/remorses/usecomputer
1•xmorse•3m ago•1 comments

Blender Optimization Leads to 2x Performance on CPU Bottlenecked Scenes

https://www.phoronix.com/news/Blender-EEVEE-2x-Perf
1•rbanffy•4m ago•0 comments

NesusFix – Zero-alloc, compile-time hardened FIX engine for sub-100ns

https://github.com/SilverstreamsAI/NexusFix
1•dalvrosa•5m ago•0 comments

Show HN: OrbStudio Native Mac app for designing animated orbs with code export

https://www.orbstudio.app/
1•okbye•6m ago•0 comments

Aesthetic Programming: A Handbook of Software Studies (2021)

https://www.openhumanitiespress.org/books/titles/aesthetic-programming/
1•the-mitr•6m ago•0 comments

LG Display starts mass-producing LTPO-like 1 Hz LCD displays for laptops

https://arstechnica.com/gadgets/2026/03/lg-display-starts-mass-producing-ltpo-like-1-hz-lcd-displ...
1•rbanffy•7m ago•0 comments

Why move antimatter by road? CERN tests a truck-ready antiproton trap

https://phys.org/news/2026-03-antimatter-road-cern-truck-ready.html
1•Brajeshwar•8m ago•0 comments

Krita 6.0 Released with Qt6 Port and Better Wayland Support

https://www.phoronix.com/news/Krita-6.0-Released
3•doener•9m ago•0 comments

Krita 5.3.0 and 6.0.0 Released

https://krita.org/en/posts/2026/krita-5.3.0-released/
2•jrepinc•9m ago•0 comments

Iran internet offline for 25 days and

https://twitter.com/netblocks/status/2036343014277165512
1•ukblewis•9m ago•1 comments

A mission NASA might kill is still returning science from Jupiter

https://arstechnica.com/space/2026/03/a-mission-nasa-might-kill-is-still-returning-fascinating-sc...
1•rbanffy•10m ago•0 comments

Guile Knots

https://www.cbaines.net/posts/announcing_guile_knots/
1•ibobev•11m ago•0 comments

Mendeleev's Inequality

https://www.johndcook.com/blog/2026/03/24/mendeleevs-inequality/
1•ibobev•12m ago•0 comments

Lines of Code Are Useful

https://entropicthoughts.com/lines-of-code
1•ibobev•12m ago•0 comments

Show HN: MangoWave – Open-source browser audio visualizer

https://mangowave.app/
1•lmascari•12m ago•0 comments

Understaffing as a Form of Enshittification

https://pluralistic.net/2026/03/22/nobodys-home/
2•toomuchtodo•12m ago•0 comments

io_uring, libaio performance across Linux kernels and an unexpected IOMMU trap

https://blog.ydb.tech/how-io-uring-overtook-libaio-performance-across-linux-kernels-and-an-unexpe...
3•tanelpoder•12m ago•1 comments

Ignition: NASA's Plan for the Moon [video]

https://www.youtube.com/watch?v=yIlTwwJv1Ac
1•trothamel•13m ago•1 comments

Keytrace (Keybase on ATProto)

https://keytrace.dev/blog/introducing-keytrace
1•bezbac•14m ago•0 comments

JetBrains Central: An Open System for Agentic Software Development

https://blog.jetbrains.com/blog/2026/03/24/introducing-jetbrains-central-an-open-system-for-agent...
1•vovak•14m ago•0 comments

BotmasterLabs: Powering an Era of Spam

https://castle.io/research/botmasterlabs-powering-an-era-of-spam/
1•avastel•14m ago•0 comments

Why your AI agents will turn against you

https://yoloai.dev/posts/ai-agent-threat-landscape/
2•kstenerud•14m ago•1 comments

DSL for reverse engineering using KiCad

https://github.com/ryanm101/revsch
2•ryanm101•15m ago•0 comments

AI Agentic for Banking and Financial Services: Credit Analysis, KYC in 2026

https://simplai.ai/blogs/agentic-ai-banking-financial-services-mortgage-kyc-credit-analysis/https...
1•Shanmugaraj-1•16m ago•0 comments