ubermon

4 ms·

1.
by ubermon 5d ago
not yet, but will be included in the distributed version
2.
by ubermon 5d ago
https://github.com/AntigmaLabs/ante
3.
Everything Is a Stream: runtime composability over compile-time plugins(antigma.ai)
4 points by ubermon 5d ago|5 comments
4.
by ubermon 1mo ago
the point is you can redirect all the telemetry to your own exporter with nothing leaves your machine and all the benefits.
5.
by ubermon 1mo ago
it can be opt-out, opt-in, but the telemetry needs to be there to actually improve the agent.
6.
by ubermon 2mo ago
it is more like an actual harness with a managed and pinned llama the control is inverse. Harness is compute and Model is data
7.
by ubermon 2mo ago
there are so much content and valuable stuff in the github repo you can pretty much recreated with your own agent.
8.
by ubermon 2mo ago
yes. Offline mode is choice, should be able to run frontier one first and then figure out how do incorporate local models as real workhorse
9.
by ubermon 2mo ago
there are still many bad players in the industry, i wouldn't mind sharing it with trusted group. But I am not yet strong enough with deal and handle all those yet.
10.
by ubermon 2mo ago
generally the challenge now is that 1. how to deal with PR spams by AI bots 2. how to make the project sustainable especially when one has no distribution. when anyone can insta remix and re-package and re-sell your hard work. For knowledg
11.
by ubermon 2mo ago
called out the most asked questions - where is the source - telemetry opt-in/opt-out in README of https://github.com/AntigmaLabs/ante
12.
by ubermon 2mo ago
I wonder my self, was notified by a friend, but I am grateful.. Probably because of the Meta's open model release?
13.
by ubermon 2mo ago
my view on harness is that it is to capturing the structural mechanism with llm interacting the world. They are a dynamic duo evolving together. The technical depth will continue to grow (e.g. /goal being the new primitive, multi-agent
14.
by ubermon 2mo ago
I love pi and share many vision and value with it. But my view on harness is that it is to capturing the structural mechanism with llm interacting the world. They are a dynamic duo evolving together. The technical depth will continue to gro
15.
by ubermon 2mo ago
Oh no, definitely not Internet Explorer. Will figure out a way to ship with source first to address the security and concern and then figure out how to do the open source development in agentic era later, I was too carried away by the compl
16.
by ubermon 2mo ago
the public repo README is updated.
17.
by ubermon 2mo ago
there is one toy version https://github.com/AntigmaLabs/nanochat-rs README updated.
18.
by ubermon 2mo ago
even with power of AI, we are mere human and still slow. Adding this to backlog.
19.
by ubermon 2mo ago
feedback received, it was carry over from the preview dev build.
20.
by ubermon 2mo ago
for now only some of core crates is migrated, will do so progressively
21.
by ubermon 2mo ago
so it is more for being self contained and works out of box if being deployed in a bare linux environment. we tried shell out to `rg` it didn't work very well and instead spending time handling the args parsing and jugging string outp
22.
by ubermon 2mo ago
nice, good to see more contributor in this space
23.
by ubermon 2mo ago
we put it in the repo README, will add migrate more into public repo as soon as possible.
24.
by ubermon 2mo ago
yes, the goal is to perfect the `ante serve` so it is easy to build gui. We are building one internally to test the protocol version
25.
by ubermon 2mo ago
will add those soon!
26.
by ubermon 2mo ago
we have a detailed launch thread explaining and show case exactly this! https://x.com/NoCommas/status/2086835536598351955
27.
by ubermon 2mo ago
Hi HN, I'm Mohan from Antigma Labs. Ante is a coding agent that ships as one self-contained ~15MB binary: the TUI, an embedded ripgrep, local PDF/OCR, and a natively managed llama.cpp engine are all inside. No runtime dependencies
28.
Show HN: Ante, a coding agent in a single binary that runs offline(github.com)
169 points by ubermon 2mo ago|93 comments
29.
by ubermon 2mo ago
yeah, that bench has a lot of room to improve, but it is the best we can find now
30.
DeepSeek V4 Flash 0731: 82.7% on Terminal-Bench 2.1 with a public harness(antigma.ai)
37 points by ubermon 2mo ago|6 comments

More ›