Show HN: A registry of agent benchmarks (including many OSS agent trajectories) https://bit.ly/4guknjO

Show HN: A registry of agent benchmarks (including many OSS agent trajectories) If you're interested in exploring what LLM-based agent systems these days actually do to solve certain benchmarks such as SWEBench or WebArena, we created a small leaderboard with our team, that allows to view a lot of public and OSS agent results including all the runtime traces (the step-by-step reasoning behind the scenes). Looking at traces is actually quite interesting, as they reveal a lot about the inner working and shortcomings of current agent system, e.g. see https://bit.ly/4gplkdl... for an example trace. https://bit.ly/4grYPVa December 23, 2024 at 09:57AM

Show HN: A simple telegram file downloader https://bit.ly/4gLNUp5

Show HN: A simple telegram file downloader https://bit.ly/41IkALS December 23, 2024 at 08:54AM

Show HN: I built this website about Sikh History and don't know how code works https://bit.ly/3DtyCXI

Show HN: I built this website about Sikh History and don't know how code works I've been learning about Sikhism and Sikh history recently and, despite having Game of Thrones level drama, I found the resources really lacking and nowhere piecing it all together. I work in a developer adjacent role (ok, I'm a Product Manager) but despite working with software engineers every day I don't really get coding. I see a lot of stuff online about the death of software engineers and wanted to challenge myself to see if I could create something myself. I've been using the free tier of Anthropic's Claude AI, deployed on the free Vercel tier, spent $10 on the domain but not a penny more on anything else. It's super basic but felt good to make something myself and I learned a lot. I'd brick it at the idea of adding anything complex (or even being asked how it all works together) so I'm sure developers are safe for a while yet! https://bit.ly/3ZSY6VZ December 23, 2024 at 06:01AM

Show HN: Ephemeral VMs in 1 Microsecond https://bit.ly/41HnS1X

Show HN: Ephemeral VMs in 1 Microsecond https://bit.ly/4072FgM December 20, 2024 at 11:43AM

Show HN: Npflared serveless private NPM registry that you can host for free https://bit.ly/4gmynMy

Show HN: Npflared serveless private NPM registry that you can host for free Free and open source, npflared is a serveless private npm registry that you can self-host in order to distribute private packages for you and your team https://bit.ly/4gEqTEE December 22, 2024 at 02:16PM

Show HN: Cudair – live-reloading for developing CUDA applications https://bit.ly/4gj4T29

Show HN: Cudair – live-reloading for developing CUDA applications cudair enable live-reloading for developing CUDA applications like golang-air. https://bit.ly/3Pb1x5o December 22, 2024 at 09:00AM

Show HN: GitHub-assistant – Natural language questions from your GitHub data https://bit.ly/3BTyfoG

Show HN: GitHub-assistant – Natural language questions from your GitHub data Simon(sfarshid) and I spend a lot of time on GitHub. As data nerds we put together a quick tool to explore your repository’s data. How it works: - Data Loading: We use dlt to pull data (issues, PRs, commits, stars) from GitHub - Semantic Layer: Relta wraps the underlying dataset into a semantic layer so the LLM doesn’t hallucinate. - Text-to-SQL: A text-to-SQL agent transforms your plain-English question into a query using the semantic layer - Generative Charts: assistant-ui dynamically generates a chart based on the SQL query - Refinements: If the semantic layer can’t handle your question, our agent submits semantic layer improvements via pull requests Hosted version: https://bit.ly/4grR2qr Demo Video: https://youtu.be/ATaf98nID5c Check out the repo + hosted version and let us know what you think. https://bit.ly/49M4tit December 22, 2024 at 01:41AM

Show HN: Web worker and polling exp / HN client app https://bit.ly/3ZRPoam

Show HN: Web worker and polling exp / HN client app demo and src inside newstories will be loaded in web worker polling for comments refresh in main thread configurable worker cache https://bit.ly/3BxrOrw December 21, 2024 at 08:58PM

Show HN: Eonfall – A new third-person co-op action game built for the web https://bit.ly/3Wf2wWj

Show HN: Eonfall – A new third-person co-op action game built for the web Hi all, I'm excited to share Eonfall with Hacker News Community! It's been 2-years in the making built by a 2 man team. Eonfall, is a new third-person co-op action game with rogue-lite elements built exclusively for the web! We've finally reached a release candidate state and set our official public release date for Jan 15th! The game's current version 5.0.0-beta is live and available to test play today! Unity game engine was used to develop the game along with other services to handle the backend, and Nuxt 3 + Nuxt UI to handle the front-end. We welcome any and all questions, feedback & suggestions! Thanks all, Jon https://bit.ly/3VQ280b December 21, 2024 at 05:45PM

Show HN: City Summit – buildings data visualization project https://bit.ly/3DrG8lT

Show HN: City Summit – buildings data visualization project https://bit.ly/4iK003N December 21, 2024 at 08:49AM

Show HN:Free Online Tool to Experience Microsoft's MarkItdown https://bit.ly/4gJFKNL

Show HN:Free Online Tool to Experience Microsoft's MarkItdown https://bit.ly/4iLIADW December 21, 2024 at 06:13AM

Show HN: Find Domains Fast (Prices, SEO, AI and Alerts) https://bit.ly/49N2IkR

Show HN: Find Domains Fast (Prices, SEO, AI and Alerts) After being frustrated for years with existing solutions, I'm working on a new domain finder. The idea is to combine everything I was missing: instant speed, good design, multiple extensions (with customizable search) and hack domains, real-time price comparison, SEO insights, domain name generation, meaning and sentiment analysis, favorites saving, custom alerts, management of all my domains all from one place. If you try it, I'd love to read about it. https://bit.ly/3ZL9R0r December 20, 2024 at 11:46AM

Show HN: Interactive graphs in Rerun with a Rust port of D3-force https://bit.ly/4gmZZRD

Show HN: Interactive graphs in Rerun with a Rust port of D3-force Rerun 0.21 comes with a new graph viewer that's written in Rust and runs in the browser via wasm. It's powered by a new force based layout engine that is a port of much of d3-force to Rust. (The release also contains some other cool stuff like undo/redo implemented on top of a timeseries DB.) We built this with applications in robotics and spatial computing in mind but would love to hear feedback from folks that would see this as useful in other domains as well. https://bit.ly/422HI8b December 20, 2024 at 10:42AM

Show HN: Animated Wallpaper – Reacts to Mouse https://bit.ly/49MRMUm

Show HN: Animated Wallpaper – Reacts to Mouse https://bit.ly/4gNrbch December 19, 2024 at 11:34PM

Show HN: SEO Keyword Grouping and SERP Checkers https://bit.ly/4foV818

Show HN: SEO Keyword Grouping and SERP Checkers https://bit.ly/3VMbppL December 19, 2024 at 12:59AM

Show HN: Yakari – Interactive TUIs for CLI Tools https://bit.ly/3P3UUBH

Show HN: Yakari – Interactive TUIs for CLI Tools Hi HN! I wanted to share Yakari, a tool I built to make command-line interfaces more approachable through interactive TUIs. If you've ever forgotten CLI flags or needed to look up command syntax, this might help. Yakari turns complex commands into interactive menus. Users can navigate through options with simple key presses instead of memorizing complex command structures. If you've used Emacs and Magit (or any other Transient) before, the interface will feel familiar. Features: - Transform CLIs into guided menus - Create custom menus for any CLI - Support for flags, named parameters, choices, and interactive inputs - Command history and contextual help You can try it out without installing thanks to uv [1]: uvx --from yakari ykr demo # Play with a demo showcasing different argument types uvx --from yakari ykr git # Try the git menu in any git repo The project is built with Python using Textual and is heavily inspired by Emacs' Transient. I'd love feedback from both CLI users and developers. What tools would you find most useful to have menus for? How could this make your terminal workflows easier? [1] https://bit.ly/4iKnK80 https://bit.ly/4iGhyOh December 18, 2024 at 11:54PM

Show HN: HearTomo – an app that understands your heart https://bit.ly/4glFbdj

Show HN: HearTomo – an app that understands your heart https://apple.co/4fniUKL December 18, 2024 at 04:15AM

Show HN: An Open Source Equilizer Plugin https://bit.ly/3BHPeKA

Show HN: An Open Source Equilizer Plugin a5eq.lv2 is a versatile LV2 plugin featuring a high-performance 5-band equalizer, equipped with a Low Shelf, three Peaking Filters, and a High Shelf. The goal of a5eq.lv2 is to deliver optimized performance on both AMD64 and ARM64 architectures. By leveraging SIMD instructions, the plugin ensures efficient and reliable operation. As the author of a5eq.lv2, I welcome feature requests and ideas for improvement. https://bit.ly/41BSfXp December 18, 2024 at 02:35AM

Show HN: Play the prototype of our RPG Game, where you can ASCEND to the ASTRAL https://bit.ly/4gjbKIP

Show HN: Play the prototype of our RPG Game, where you can ASCEND to the ASTRAL https://bit.ly/4fnk1dI December 17, 2024 at 09:22PM

Show HN: A simple air quality app https://bit.ly/3ZWvdJH

Show HN: A simple air quality app I made this app for myself, but it's probably worth sharing. When I tried to find a nice app that I could use to check air quality multiple times a day, I was shocked. All of them looked boring with traffic-light coding and overcrowded with unnecessary data. On top of that, they either contained ads or required payment. I didn't want to use any of those apps every day, so I thought, "Well, how hard could it be?" and accepted the challenge. I expected to finish development in a month, but since I'm a designer and don't normally program apps, I had to learn Swift on the go. This required a month more than I expected. In my designs, I'm usually way too focused on efficiency. Since this was my personal project, I wanted to experiment, so I didn't use standard controls or color coding. Inspired by avant-garde style, I used forms and movement to convey air quality, as well as an asymmetrical layout and overflowing visualizations. I enjoy using the app, and I hope you will, too. https://apple.co/4iIt4ZC December 18, 2024 at 12:19AM