etienne_l
Re:Show HN: A Playable Architectural Facade Synth
on Tuesday October 13, 2026 @04:54AM (#50068661)
This is awesome! I can tell this is a huge amount of work selecting the correct type and amount of variables.
Re:Show HN: Let Claude Code answer with a diagram page instead of a wa...
on Tuesday October 13, 2026 @04:50AM (#50068629)
That's a great idea! next step, let Claude answer with comic book panels :-D
Re:HN has lost the joy for me
on Wednesday October 07, 2026 @04:41AM (#49989957)
We are spoiled
Re:AI is now capable of developing its own inference hardware
on Wednesday October 07, 2026 @04:38AM (#49989937)
You never know what will be remembered as the birth of the singularity. Could be a small github repo like this one, who knows.
Re:Decisions API is in public beta
on Wednesday October 07, 2026 @04:31AM (#49989893)
This is more a great signal marketwise than anything else. But still this is confirming the buzz.
Re:OpenScreen, the open-source Screen Studio alternative, has a new ma...
on Tuesday September 01, 2026 @08:10AM (#49520880)
I maintain this one. I was the #2 contributor on the original repo, and
Siddharth linked here from his README when he archived it.
Nine releases since June. Biggest change is the compositor: it used to run in Electron, now it's native (D3D11, Metal, Vulkan). Export on my test machine went from about 10 fps to about 126. Linux got a lot of work too, and there are in-place updates now, plus Microsoft Store / Fedora / Nix / AUR packages.
Happy to answer questions.
OpenScreen, the open-source Screen Studio alternative, has a new maintainer
Re:N8n-as-code: Stop n8n AI hallucinations with TS decorators
on Wednesday March 11, 2026 @09:19AM (#47335212)
I spent the last few weeks building an abstraction layer for n8n to fix a specific problem: LLM hallucinations in AI IDEs.
When you ask Cursor or Windsurf to build an n8n workflow, it often breaks the JSON schema. By bringing the n8n ontology into TypeScript with decorators, we provide the AI with the strict context it needs to be 10x more reliable.
It's a pure GitOps approach to automation. Curious to see if others here are moving away from 'No-Code' UI towards 'As-Code' for better AI reliability.
N8n-as-code: Stop n8n AI hallucinations with TS decorators
Re:Show HN: I am building a monitoring app to catch product issues
on Monday December 15, 2025 @09:23AM (#46274924)
Are you production ready ?
Re:Show HN: Stimm – Low-Latency Voice Agent Platform (Python/WebRTC)
on Monday December 15, 2025 @09:02AM (#46274666)
What it is: Stimm (v0.1 Public Beta) is an Open Source (AGPL v3) platform designed to orchestrate low-latency voice conversations with AI agents. We focused heavily on the Python/FastAPI event loop and media transport (using LiveKit for WebRTC) to achieve a near real-time feel.
Why we built it: To provide a robust, modular foundation for builders. Stimm handles the complex real-time pipeline (WebRTC, VAD, STT/LLM/TTS streaming) so you can focus on the AI. It includes built-in SIP telephony support and is fully Dockerized.
Feedback wanted: We're looking for technical feedback on the architecture, particularly the audio stream handling.
Looking forward to the discussion!