frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Attention Sinks and Compression Valleys in LLMs

https://arxiv.org/abs/2510.06477
1•alexkranias•30s ago•0 comments

AI Chat Evaluation of the Formal Language in He Xin's PEPC System 2

1•nikicsy•49s ago•0 comments

Hand tool rewrites ancient Egyptian history

https://www.popsci.com/science/ancient-egypt-hand-tool/
1•delichon•54s ago•0 comments

A note about personal security

https://werd.io/a-note-abo/
1•sdoering•1m ago•0 comments

AI Chat Evaluation of the Formal Language in He Xin's PEPC System

1•nikicsy•2m ago•0 comments

A Note on File History in Emacs

https://brainbaking.com/post/2026/02/a-note-on-file-history-in-emacs/
1•Brajeshwar•2m ago•0 comments

Revisionist History – Aliens, Secrets and Conspiracies

https://steveblank.com/2026/02/10/revisionist-history-aliens-secrets-and-conspiracies/
1•Brajeshwar•2m ago•0 comments

Show HN: cbt (C++ Build Tool)

https://github.com/swar-mukh/cbt
1•swar-mukh•2m ago•0 comments

Open model StepFun-3.5 is #1 on MathArena, an uncheatable math benchmark

https://twitter.com/CyouSakura/status/2021511358626554322
1•diyer22•2m ago•0 comments

Show HN: Bitcoin, GEB, and Bach's fugues share the same structural move

https://falsework.dev/
1•falsework•2m ago•0 comments

Functional Programming in M4

https://minnie.tuhs.org/pipermail/tuhs/2020-August/022108.html
1•fanf2•4m ago•0 comments

AI makes it easier to build the wrong thing faster

https://newsletter.masilotti.com/p/ai-makes-it-easier-to-build-the-wrong
1•joemasilotti•4m ago•1 comments

Show HN: I built a macOS desktop toy that patrols while you work

https://airwolfspace.com/tinytanks
1•kailuo•4m ago•0 comments

Poison at Play: Unsafe lead levels found in half of New Orleans playgrounds

https://veritenews.org/2026/02/05/poison-at-play-playgrounds-lead-levels/
1•hn_acker•4m ago•0 comments

Unresponsive Buttons on My Fastest Hardware

https://blog.jim-nielsen.com/2026/unresponsive-buttons/
2•speckx•4m ago•0 comments

AI-First Company Memos

https://the-ai-native.company/
1•bobismyuncle•5m ago•0 comments

How to Test ProxySQL Read/Write Split with Sysbench

https://rendiment.io/mysql/proxysql/2026/02/03/sysbench-proxysql.html
1•nethalo•6m ago•0 comments

The singularity won't be gentle – by Nate Silver

https://www.natesilver.net/p/the-singularity-wont-be-gentle
1•rbanffy•7m ago•0 comments

A New Computer Could Replace Electricity with Light

https://www.popularmechanics.com/science/a70223544/computer-could-replace-electricity-with-light/
1•falcor84•7m ago•0 comments

Show HN: Health.md - Apple Health → Markdown

https://healthmd.isolated.tech/
1•codybontecou•8m ago•0 comments

PicoClaw: Ultra-Efficient AI Assistant in Go

https://github.com/sipeed/picoclaw
1•wicket•9m ago•0 comments

AITools.coffee – GitHub metrics observatory tracking 27K+ open-source AI repos

https://aitools.coffee
1•alexela84•9m ago•1 comments

AI Agents 101: From Concept to Code (No Frameworks Required)

https://medium.com/@kamil.tustanowski/ai-agents-101-from-concept-to-code-no-frameworks-required-2...
1•semerkchet•9m ago•0 comments

Databases should contain their own Metadata – Use SQL Everywhere

https://floedb.ai/blog/databases-should-contain-their-own-metadata-instrumentation-in-floe
4•matheusalmeida•10m ago•0 comments

Seeking Order in Chaos

https://garrit.xyz/posts/2026-02-11-on-seeking-order-in-chaos
3•garritfra•10m ago•0 comments

Show HN: Funxy – A typed scripting language that embeds into Go apps

https://github.com/funvibe/funxy
2•funbitty•10m ago•0 comments

The jarring experience of developing today

https://its.beer/thoughts/the-jarring-experience-of-developing-today
1•beerd•11m ago•0 comments

Kiro: DeepSeek, MiniMax, and Qwen now available as open weight model options

https://kiro.dev/changelog/models/deepseek-minimax-and-qwen-now-available-as-open-weight-model-op...
2•siegers•11m ago•0 comments

Terence Tao: Why I Co-Founded SAIR

https://www.youtube.com/watch?v=Z5GKnb4H_bM
1•nyc111•13m ago•0 comments

Maia 200: The AI accelerator built for inference

https://blogs.microsoft.com/blog/2026/01/26/maia-200-the-ai-accelerator-built-for-inference/
1•MarlonPro•16m ago•0 comments
Open in hackernews

Show HN: AgentWire – Talk to your AI coding agents by voice, across machines

https://github.com/dotdevdotdev/agentwire-dev
1•prradox•1h ago

Comments

prradox•1h ago
I run multiple Claude Code and OpenCode sessions at the same time — sometimes on my laptop, sometimes on remote devboxes over SSH. The friction that bugged me most was context switching: I'd be thinking through a problem, pacing around, and then have to sit down, find the right terminal window, and type out what I wanted the agent to do. I kept wishing I could just say it out loud.

So I built AgentWire. It's a self-hosted CLI + browser portal that sits on top of tmux. You open the portal on your phone or tablet, push-to-talk, and your voice gets transcribed and routed to whichever agent session you pick. The agent talks back via TTS. It works across machines — SSH into a remote box, and those sessions show up in the same portal.

The core loop: speak a prompt, agent works on it, agent speaks the result back. You can be on the couch, on a train, wherever — as long as your portal is accessible (I use a Cloudflare tunnel).

It also does multi-agent orchestration. One session can be the "leader" that spawns worker panes, delegates tasks, and collects summaries. This is useful when you want parallel work across a codebase without the agents stepping on each other.

You can run everything on one machine, or distribute it across several like I do — I have TTS running on a Windows WSL box with a GPU, several dev boxes (local and remote) running various projects, and my main local machine running the portal and STT.

Install: pip install agentwire-dev (or uv tool install agentwire-dev). No cloud, no accounts, no sign-up. Your audio stays on your machines. TTS runs on your own GPU or serverless (RunPod), STT via Whisper.

I use this daily. The docs are thin, and I'm adding tons of QoL and better error paths daily too. But the core voice workflow and session management are solid.

Curious about:

1. Is push-to-talk the right model? Or would continuous listening / wake-word work better for coding?

2. If you run multiple AI coding agents, what would you want from a coordination layer?

3. What's the first thing that would need to improve for you to actually try this?

Python, tmux, MCP. Works with Claude Code, OpenCode, or anything that runs in a terminal.