Show HN: Text to 3D simulation on a map (does history pretty well)
10 by lukehollis | 5 comments on Hacker News.
Simulate anything on a map from a text prompt -- and conduct risk analysis against LiveUA map's global realtime data points from social media and news sources. I trained a GPT-2-size model on historical incident data used to predict things that will go wrong. As historian Benjamin Breen mentions, the leading language models are good historians, so the application will simulate historical events pretty well also. I include a Multi-Agent RL Urban Mobility model in progress displayed on the map as small white cubes representing traffic and pedestrians. Around SF, it uses real census data and other sources for semantically meaningful day plans, etc. It will populate where you move the map, albeit a little slowly. This is based on previous work on my GitHub--I hope to connect it to Unreal's city samples project soon. The simulations are pretty simple so far but will grow in complexity soon. I won the AGI House World Models Hackathon with this and the MARL model. Many thanks to Shota Matsuda and Garrett Johnson for the cloud and atmospheric effects libraries on Github at takram-design-engineering/three-geospatial Glad for feedback, and thanks for trying it out!
10 by lukehollis | 5 comments on Hacker News.
Simulate anything on a map from a text prompt -- and conduct risk analysis against LiveUA map's global realtime data points from social media and news sources. I trained a GPT-2-size model on historical incident data used to predict things that will go wrong. As historian Benjamin Breen mentions, the leading language models are good historians, so the application will simulate historical events pretty well also. I include a Multi-Agent RL Urban Mobility model in progress displayed on the map as small white cubes representing traffic and pedestrians. Around SF, it uses real census data and other sources for semantically meaningful day plans, etc. It will populate where you move the map, albeit a little slowly. This is based on previous work on my GitHub--I hope to connect it to Unreal's city samples project soon. The simulations are pretty simple so far but will grow in complexity soon. I won the AGI House World Models Hackathon with this and the MARL model. Many thanks to Shota Matsuda and Garrett Johnson for the cloud and atmospheric effects libraries on Github at takram-design-engineering/three-geospatial Glad for feedback, and thanks for trying it out!
CRLF Injection in `–proxy-header` allows extra HTTP headers (CWE-93)
5 by oblivionsage | 2 comments on Hacker News.
5 by oblivionsage | 2 comments on Hacker News.
The Lisp in the Cellar: Dependent Types That Live Upstairs [pdf]
12 by todsacerdoti | 0 comments on Hacker News.
12 by todsacerdoti | 0 comments on Hacker News.
Show HN: Astra – a new js2exe compiler
12 by qwertycodepl | 1 comments on Hacker News.
Hi everyone I'm new here and i wanted to introduce my project i've been working on.Astra is a simple but powerful node.js to exe compiler. It uses esbuild and Node SEA. It uses postject to inject your code to nodejs binary. It focuses more on compiling cli and Servers like pkg or nexe (express) than fullstack applications like electron or tauri. It has rich ESM and typescript support. It has good DX and cli UX. I made it bc i didn't like using pkg or nexe, they cause a lot of problems with esm. LIMITATIONS: Now it has problems compiling projects with depencides containing binaries (e.g. bcrypt, rcedit), and it compiles only for Windows but i'm working on it If you like it, leave a and comment what you think about it!
12 by qwertycodepl | 1 comments on Hacker News.
Hi everyone I'm new here and i wanted to introduce my project i've been working on.Astra is a simple but powerful node.js to exe compiler. It uses esbuild and Node SEA. It uses postject to inject your code to nodejs binary. It focuses more on compiling cli and Servers like pkg or nexe (express) than fullstack applications like electron or tauri. It has rich ESM and typescript support. It has good DX and cli UX. I made it bc i didn't like using pkg or nexe, they cause a lot of problems with esm. LIMITATIONS: Now it has problems compiling projects with depencides containing binaries (e.g. bcrypt, rcedit), and it compiles only for Windows but i'm working on it If you like it, leave a and comment what you think about it!
Show HN: A Tiling Window Manager for Windows, Written in Janet
13 by agentkilo | 1 comments on Hacker News.
Hi HN! I read[1] about Janet[2] some time ago, then immediately got impressed by the enthusiasm of its community, and by the language itself, so I started playing with it. At the time I was searching for a tiling window manager for Windows, and unavoidably the idea of scratching my own itch with Janet got hold of me, so Jwno was born. Simply put, Jwno is a keyboard-driven tiling window manager for Windows, scriptable with Janet. But since it has a complete Lisp runtime, and a thin wrapper library for Win32 APIs[3], you can certainly do much more with it. I hope you'll enjoy playing with it as much as I enjoyed building it. And yes, I use StumpWM on the Linux side, by the way. [1]: https://ift.tt/bwjCJm2 [2]: https://janet-lang.org/ [3]: https://ift.tt/1OS9vMV
13 by agentkilo | 1 comments on Hacker News.
Hi HN! I read[1] about Janet[2] some time ago, then immediately got impressed by the enthusiasm of its community, and by the language itself, so I started playing with it. At the time I was searching for a tiling window manager for Windows, and unavoidably the idea of scratching my own itch with Janet got hold of me, so Jwno was born. Simply put, Jwno is a keyboard-driven tiling window manager for Windows, scriptable with Janet. But since it has a complete Lisp runtime, and a thin wrapper library for Win32 APIs[3], you can certainly do much more with it. I hope you'll enjoy playing with it as much as I enjoyed building it. And yes, I use StumpWM on the Linux side, by the way. [1]: https://ift.tt/bwjCJm2 [2]: https://janet-lang.org/ [3]: https://ift.tt/1OS9vMV
If an AI agent can't figure out how your API works, neither can your users
87 by mattmarcus | 38 comments on Hacker News.
87 by mattmarcus | 38 comments on Hacker News.
Launch HN: Opusense (YC X25) – AI assistant for construction inspectors on site
11 by rcody | 1 comments on Hacker News.
Hi HN, we're Roya and Michael, co-founders of Opusense AI ( https://ift.tt/fpauHo6 ), a tool to help engineers and consultants automatically generate construction site reports from typed or voice notes, plus photos. Here’s a video: https://www.youtube.com/watch?v=u3Pi1iih1_Y . Before this, I (Roya) worked in human-machine interaction at Huawei, and before that as a construction site inspector for civil engineering firms. I have a PhD in Civil Engineering, and in my experience reporting was by far the most tedious and mind-numbing part of the job. You’d walk around a site all day taking short notes (maybe, often you'd rely on memory) and snapping photos, then go to three more sites before finally making it back to the office and try to remember everything you wanted to write. Sometimes you’d fill in gaps from memory or you’d keep it purposefully vague. Reports had to be consistent, branded, and checked by senior engineers. It was a huge time sink across the team. Writing reports was the worst part of the job, so we built Opusense to get rid of it. On-site, users type or dictate short notes (e.g. “rebar exposed east end of slab”), and the tool turns them into full sentences, paragraphs, tables, or photo captions in a report template that matches the firm’s format. You can work offline, and it syncs automatically when back online. Most inspection and reporting tools are built for checklist-style workflows (which is great for home inspections or punch lists), but civil, structural, environmental, or geotechnical engineers usually need freeform notes, not radio buttons. This is a particularly good fit for LLMs because engineering field reports live in a constrained, conventional domain: similar language, repeated structures, and highly standardized content across firms and projects. There’s a lot of redundancy and grunt work, summarizing the same site conditions, formatting repetitive data, translating field notes into polished paragraphs, all of which LLMs handle well with the right prompting and guardrails. We’re not generating arbitrary prose; we’re transforming structured inputs (notes, images, forms) into structured outputs, with firm-defined templates and required fields that minimize the risk of hallucination. When facts matter (e.g. test results or measurements), we keep them grounded in the user’s input, the model doesn’t invent data because there’s nothing for it to invent. This makes it one of those cases where LLMs aren’t just a novelty, they're genuinely the best tool for the job. Under the hood, we use a combination of prompt-engineered LLMs and firm-specific formatting rules to get outputs that don’t just sound good, but also look right. We’ve recently added translation features, and we’re iterating quickly based on field feedback. We charge per seat and are deployed at mid size firms, and trialing with some multinational engineering firms who have thousands of reports to file each week. We're also starting to see interest from construction managers and developers who do their own internal QA reporting. We don't have a self-serve way to try out the product yet, because the way our business works requires templates to be customized by company. But there’s a demo at https://www.youtube.com/watch?v=u3Pi1iih1_Y , and if you want to poke around the UI yourself, here’s a sample account to log in with: login: hndemo@opusense.com password: OpusenseHacker2025 The app is available for download on the Apple and Google Play stores. When sample reports are generated, you can log into the web interface to also view them online through our website (www.opusense.com) with the same login credentials. We’d love to hear how others are thinking about tools for field work, reporting, or similar workflows (engineering, architectural, etc.). If you’ve built in this space, or have thoughts on how to improve it, we’re all ears!
11 by rcody | 1 comments on Hacker News.
Hi HN, we're Roya and Michael, co-founders of Opusense AI ( https://ift.tt/fpauHo6 ), a tool to help engineers and consultants automatically generate construction site reports from typed or voice notes, plus photos. Here’s a video: https://www.youtube.com/watch?v=u3Pi1iih1_Y . Before this, I (Roya) worked in human-machine interaction at Huawei, and before that as a construction site inspector for civil engineering firms. I have a PhD in Civil Engineering, and in my experience reporting was by far the most tedious and mind-numbing part of the job. You’d walk around a site all day taking short notes (maybe, often you'd rely on memory) and snapping photos, then go to three more sites before finally making it back to the office and try to remember everything you wanted to write. Sometimes you’d fill in gaps from memory or you’d keep it purposefully vague. Reports had to be consistent, branded, and checked by senior engineers. It was a huge time sink across the team. Writing reports was the worst part of the job, so we built Opusense to get rid of it. On-site, users type or dictate short notes (e.g. “rebar exposed east end of slab”), and the tool turns them into full sentences, paragraphs, tables, or photo captions in a report template that matches the firm’s format. You can work offline, and it syncs automatically when back online. Most inspection and reporting tools are built for checklist-style workflows (which is great for home inspections or punch lists), but civil, structural, environmental, or geotechnical engineers usually need freeform notes, not radio buttons. This is a particularly good fit for LLMs because engineering field reports live in a constrained, conventional domain: similar language, repeated structures, and highly standardized content across firms and projects. There’s a lot of redundancy and grunt work, summarizing the same site conditions, formatting repetitive data, translating field notes into polished paragraphs, all of which LLMs handle well with the right prompting and guardrails. We’re not generating arbitrary prose; we’re transforming structured inputs (notes, images, forms) into structured outputs, with firm-defined templates and required fields that minimize the risk of hallucination. When facts matter (e.g. test results or measurements), we keep them grounded in the user’s input, the model doesn’t invent data because there’s nothing for it to invent. This makes it one of those cases where LLMs aren’t just a novelty, they're genuinely the best tool for the job. Under the hood, we use a combination of prompt-engineered LLMs and firm-specific formatting rules to get outputs that don’t just sound good, but also look right. We’ve recently added translation features, and we’re iterating quickly based on field feedback. We charge per seat and are deployed at mid size firms, and trialing with some multinational engineering firms who have thousands of reports to file each week. We're also starting to see interest from construction managers and developers who do their own internal QA reporting. We don't have a self-serve way to try out the product yet, because the way our business works requires templates to be customized by company. But there’s a demo at https://www.youtube.com/watch?v=u3Pi1iih1_Y , and if you want to poke around the UI yourself, here’s a sample account to log in with: login: hndemo@opusense.com password: OpusenseHacker2025 The app is available for download on the Apple and Google Play stores. When sample reports are generated, you can log into the web interface to also view them online through our website (www.opusense.com) with the same login credentials. We’d love to hear how others are thinking about tools for field work, reporting, or similar workflows (engineering, architectural, etc.). If you’ve built in this space, or have thoughts on how to improve it, we’re all ears!
27000 Dragons and 10'000 Lights: GPU-Driven Clustered Forward Renderer
14 by logdahl | 1 comments on Hacker News.
14 by logdahl | 1 comments on Hacker News.
Robin: A multi-agent system for automating scientific discovery
20 by nopinsight | 4 comments on Hacker News.
20 by nopinsight | 4 comments on Hacker News.
Show HN: Juvio – UV Kernel for Jupyter
23 by okost1 | 9 comments on Hacker News.
Juvio brings inline, PEP 723-style dependency management and automatic, ephemeral env setup to Jupyter notebooks.
23 by okost1 | 9 comments on Hacker News.
Juvio brings inline, PEP 723-style dependency management and automatic, ephemeral env setup to Jupyter notebooks.