Show HN: ArchGW – An open-source intelligent proxy server for prompts https://bit.ly/3XssNRu

Show HN: ArchGW – An open-source intelligent proxy server for prompts Hi HN! This is Salman, Adil, Shuguang and Co working on ArchGW[1] - an open-source lightweight proxy server for prompts - written in Rust and built on top of Envoy[2]. Arch moves the critical but pesky handling and processing of prompts: task understanding, prompt routing, safety, and observability - outside business logic. Its an edge and egress proxy for agentic apps. We've talked to 100s of developers at places like Twilio, GE Healthcare, Redhat, Square, etc and there was a consistent theme in building AI apps: to move past a nascent demo they are left to their own devices in building out middle ware capabilities so that developers can move faster and ship with confidence. Today, the approach to building an enterprise-ready AI app is cobbling together a large set of mono-functional tools, adding LLM-based preprocessing steps to determine safety (e.g. applying governance and guardrails), ask clarifying questions to improve task performance, support common agentic operations by packaging and managing function calling scenarios manually, etc. Not to mention, all the undifferentiated work in incorporating different LLM models and versions, and managing resiliency, retries and fallback logic. ArchGW was built with the belief that prompts are nuanced and opaque user requests, which require the same capabilities as traditional HTTP requests including secure handling, intelligent routing, robust observability, and integration with backend (API) systems for personalization – outside business logic. We help built Envoy while at Lyft and think its offers a great foundation to build a proxy to manage traffic for prompts. Here are some additional details about the open source project. ArchGW is written in rust, and the request path has three main parts: * Listener subsystem which handles downstream (ingress) and upstream (egress) request processing. * Prompt handler subsystem. This is where ArchGW makes decisions on the safety of the incoming request via its prompt_guard primitive and identifies where to forward the conversation to via its prompt_target primitive. * Model serving subsystem is the interface that hosts all the lightweight LLMs[3] engineered in ArchGW and offers a framework for things like hallucination detection of our these models We loved building this open source project, and our belief is that this infrastructure primitive would help developers build faster, safer and more personalized agents without all the manual prompt engineering and systems integration work needed to get there. We hope to invite other developers to use and improve Arch. Please give it a shot and leave feedback here, or at our discord channel [4] Also here is a quick demo of the project in action [5]. You can check out our public docs here at [6]. Our models are also available here [7]. [1] https://bit.ly/48UhplX [2] https://bit.ly/3rJ29qp [3] https://bit.ly/3OjXJOR ... [4] https://bit.ly/3OiQAOA ... [5] https://www.youtube.com/watch?v=I4Lbhr-NNXk [6] https://bit.ly/4fFT19S [7] https://bit.ly/3OiwZhM https://bit.ly/48UhplX March 4, 2025 at 10:14PM

Show HN: Scholium, Your Own Research Assistant https://bit.ly/4hVoSVr

Show HN: Scholium, Your Own Research Assistant I built an AI-powered research agent designed to efficiently discover, summarize, and cite relevant academic papers based on user queries. As a university student, I've written my share of essays and have also served as a copy editor for our student newspaper. During fact-checking, I noticed that Google often prioritizes unreliable and unscholarly resources—such as Medium articles, Reddit posts, and LinkedIn content—in its top results over scholarly ones. For instance, searching "Transformers" yields six blogs and articles before finally listing the Vaswani (2017) paper. This makes gathering credible sources and verifying facts tedious and time-consuming. I realized that much of the repetitive work involved in fact-checking and source collection could be streamlined using a vector database paired with a retrieval model, inspiring me to create Scholium, an AI-driven research assistant that recommends and summarizes academic papers relevant to your queries. Currently, Scholium has access to all papers on arXiv, and my plan is to make Scholium into a search engine for research, kinda like a Google or Perplexity for papers. Please check out the repository, give it a star, and let me know your thoughts—I would greatly appreciate your feedback! Web App: https://bit.ly/4bpyXHF Repo: https://bit.ly/4klB1o4 https://bit.ly/4klB1o4 March 5, 2025 at 12:51AM

Show HN: FlakeUI https://bit.ly/41PYU09

Show HN: FlakeUI https://bit.ly/4h25tkx March 3, 2025 at 06:29AM

Show HN: Puffin Tools – Free WebAssembly tools in the browser https://bit.ly/4bntJMI

Show HN: Puffin Tools – Free WebAssembly tools in the browser Hi! I wanted to share my little "adventure" that I had in the past ~2 weeks. Basically, I got tired of constant thinking and I decided to build something, anything. By accident I learned about thing called WebAssembly (basically you can execute pretty complicated programs locally in the browser without sending anything anywhere) so I immediately assumed that it could be used for a set of small, private tools. Usually I am building with Python, Django, PostgreSQL, hosting it on PythonAnywhere. So this time I decided that I will step out of my comfort zone even further! So I went with: - Rust (for tools that will be compiled to WebAssembly) - Zola (static site generator since I don't need anything "dynamic") - AWS S3 + AWS CloudFront (for hosting this static website) (I even used AWS to generate my SSL certificate) Let me know what you think, and of course if I can already sell my website for million dollars :D Cheers! https://bit.ly/4h3qVW9 March 3, 2025 at 08:04AM

Show HN: Free Kindle Scribe Weekly Planner Creator https://bit.ly/4bqukx8

Show HN: Free Kindle Scribe Weekly Planner Creator https://bit.ly/4bqukNE February 28, 2025 at 02:12PM

Show HN: Prompting LLMs in Bash scripts https://bit.ly/4kq7Hgf

Show HN: Prompting LLMs in Bash scripts https://bit.ly/4bqNYsF https://bit.ly/4bpFZMH February 27, 2025 at 08:46PM

Show HN: Image comparison slider in 6 lines of JavaScript https://bit.ly/41h0ypU

Show HN: Image comparison slider in 6 lines of JavaScript https://bit.ly/4imRnva March 2, 2025 at 11:41PM

Show HN: Robyn – "Batman Inspired" Python Web Framework Built with Rust https://bit.ly/41p8eGA

Show HN: Robyn – "Batman Inspired" Python Web Framework Built with Rust https://bit.ly/41pmpvc March 2, 2025 at 08:56AM

Show HN: I built a memory-safe web server in Rust (currently in beta) https://bit.ly/4h7ca4y

Show HN: I built a memory-safe web server in Rust (currently in beta) https://bit.ly/4h3lgzs March 2, 2025 at 08:35AM

Show HN: What did you do last week? – Evaluates your 5 bullet points https://bit.ly/4kjaGqX

Show HN: What did you do last week? – Evaluates your 5 bullet points https://bit.ly/4bsFP79 March 2, 2025 at 04:57AM

Show HN: World-Price – Making international pricing transparent and reassuring https://bit.ly/43hA1LE

Show HN: World-Price – Making international pricing transparent and reassuring As someone living outside a country with a popular currency, I hated wanting a product but not being able to buy it due to a different currency. Either it's not allowed or there are immense fees. So I built World-Price, an embeddable pricing table generator that integrates with stripe to generate clean prices (either .99 prices or .00 prices) for every currency to be able to accept every user. Right now I have released a somewhat MVP, and would love your feedback if this provides use to you. Enjoy! https://bit.ly/3Xo6Oey March 2, 2025 at 12:39AM

Show HN: Simplifying Backend Testing with qapir.io https://bit.ly/3XmXxmI

Show HN: Simplifying Backend Testing with qapir.io Backend and API testing can be tedious—especially when dealing with multi-step workflows, deep validation of API responses, and complex edge cases. We've built qapir.io to make this easier, offering a no-code solution for defining and running backend and API tests without writing any code. What It Offers: * No-Code Test Creation – Define backend tests with an intuitive YAML-based syntax. * Support for Complex Scenarios – Easily test multi-step API workflows and chained API calls. * Human-Readable Reporting – Clear, structured test reports that make debugging easier. * Support for multiple protocols - Currently supports plain HTTP and GraphQL, planning to support: SQL, Redis, Kafka, etc. * An actively developing set of features – Upcoming features: Dashboard with Test-Results, Configurable HTTP-Mocks, and a solution for receiving and validating webhooks - Webhook-Interceptor Why Use It: If you need a way to test backend services without maintaining a ton of code, qapir.io provides a simple, structured approach. It’s designed to handle everything from basic API checks to more advanced multi-step scenarios—all while keeping tests readable and easy to maintain. Download it for free at https://bit.ly/3XnIbP2 , and start writing your tests in minutes! https://bit.ly/3XnIbP2 February 26, 2025 at 02:10AM

Show HN: Start GPT threads in ChatGPT (& other sites) https://bit.ly/41Cm9KO

Show HN: Start GPT threads in ChatGPT (& other sites) I noticed current LLM chat apps don't support threads for follow questions, but instead they're one long conversation. Often when I'm digging into a topic and GPT responds with a book, I want to ask follow-ups in "side chats" that I'd prefer to be inline and in context rather than jumping to bottom of the page, so I created GPT threads. It's a chrome extension where you can highlight some text and start a chat focusing on that text and incorporating surrounding context. You can have multiple threads on a page, and the extension has configurability like surrounding context size, response lengths, chat box positioning. I've found it useful for other sites too to quickly ask questions about news articles, docs, or HN threads! Downloadable in Chrome web store: https://bit.ly/3XLSblt... https://bit.ly/41AYWsl March 1, 2025 at 03:01AM

Show HN: a Rust library for creating hierarchical state machines https://bit.ly/3Xn5oRm

Show HN: a Rust library for creating hierarchical state machines I've come to prefer modeling state machines with nested states, but implementing them by hand requires a large amount of error-prone boilerplate that obscures the important logic of the state machine. I found that the existing offerings in the Rust ecosystem - while inspiring - lacked a number of features that I was looking for, so I spent the last few months building moku. Feedback is welcome! https://bit.ly/41jRmkM March 1, 2025 at 12:30AM

Show HN: Torii – a framework agnostic authentication library for Rust https://bit.ly/41f7hRc

Show HN: Torii – a framework agnostic authentication library for Rust https://bit.ly/43gwVb2 March 1, 2025 at 12:16AM

Show HN: Multi-View Stereo Software for Rapid 3D Model Generation from 2D Images https://bit.ly/3DnbBpv

Show HN: Multi-View Stereo Software for Rapid 3D Model Generation from 2D Images cuda-multi-view-stereo (CUMVS) is a C++/CUDA library for Multi-View Stereo. This project currently supports depth map estimation using PatchMatch algorithms. CUMVS achieves a speed increase of more than 5x compared to ACMM, a leading implementation of PatchMatch MVS. Here is the demo movie: https://youtu.be/g9v9b5uP68I https://bit.ly/41bzus8 February 28, 2025 at 01:16AM

Show HN: Hexle: A powerful way to search for papers and connect your ideas https://bit.ly/4ijEDFy

Show HN: Hexle: A powerful way to search for papers and connect your ideas Hi all, I find literature searches to be slow and despite all the tools nowadays, it is difficult to map out vast areas of research and then connect my ideas to what I’m trying to do (e.g. writing a review, learning something new, planning a new project, assignments, etc). Some of my peers and teachers also found this to be a problem so I built this MVP app to help people not only leverage AI search capabilities but also go on to connect their ideas, map out research projects, cite various results and organise their “stream of consciousness” as they work. I’d love to get some feedback on it - please feel free to post it either in this thread or email me at hexleteam@gmail.com. Many thanks from Sydney, Australia :) https://bit.ly/4iivOvx February 27, 2025 at 11:46PM

Show HN: SuperMassive – Fast durable, in-memory, distributed key-value database https://bit.ly/4kiiyZx

Show HN: SuperMassive – Fast durable, in-memory, distributed key-value database Hey hn! I hope you're all doing well. I’d like to share a new open-source database I’ve designed and written called SuperMassive. SuperMassive is a scalable, distributed, in-memory key-value database designed from the ground up to allow for high concurrency, fast writes, fault tolerance, and durability while remaining simple to use and efficiently scalable. The idea for SuperMassive comes from its name.. I wanted to build a key-value database that can scale infinitely, remain durable, be self-healing, consistent, and blazing fast. I also wanted to simplify sharding and replication compared to existing distributed KV stores and their protocols. SuperMassive is built to be simple yet powerful. It consists of just one multiplatform supported binary that can run in multiple modes(cluster, primary node, or replica). Nodes function as shards of your data. The design emphasizes minimal configuration, high performance, and automatic failover. While SuperMassive is still in its early stages (v1.0.2b), I’d love to hear your thoughts on the design, its architecture, and any feedback you may have :D Features~~ ~ Highly scalable Scale horizontally with ease. Simply add more nodes to the cluster. ~ Distributed Data is distributed across multiple nodes in a sharded fashion. ~ Robust Health Checking System Health checks are performed on all nodes, if any node is marked unhealthy we will try to recover it. ~ Smart Data Distribution uses a sequence-based round-robin approach for distributing writes across primary nodes. This ensures that all primary nodes get an equal share of writes. ~ Automatic Fail-over Automatic fail-over of primary nodes on write failure. If a primary node is unavailable for a write, we go to the next available primary node. ~ Parallel Read Operations Read operations are performed in parallel. ~ Consistency Management Timestamp-based version control to handle conflicts. The most recent value is always returned, the rest are deleted. ~ Fault-tolerant Replication and fail-over are supported. If a node goes down, the cluster will continue to function. ~ Self-healing Automatic data recovery. A node can recover from a journal. A node replica can recover from a primary node via a check point like algorithm. ~ Cluster Authentication with basic like auth. ~ Simple Protocol Simple protocol PUT, GET, DEL, INCR, DECR, REGX, STAT, RCNF, PING. ~ Async Node Journal Operations are written to a journal asynchronously. This allows for fast writes and recovery. ~ Multi-platform Linux, Windows, MacOS https://bit.ly/41CnAJq https://bit.ly/41z0enW - Alex https://bit.ly/41z0enW February 27, 2025 at 09:31AM

Show HN: Libredesk – Open-source customer support desk. Single binary app https://bit.ly/4gVaDyJ

Show HN: Libredesk – Open-source customer support desk. Single binary app Libredesk is a 100% free and open-source customer support desk, the backend is written in Go and the frontend is in Vue JS with ShadnCN for UI components. Unlike many "open-core" alternatives that lock essential features behind enterprise plans, Libredesk is fully open-source and plans to always stay this way. It's in alpha (v0.1.0) right now, but there’s a working demo available. I built this because I wanted a truly open and self-hosted alternative to platforms like Chatwoot, Intercom, and Zendesk. Would love feedback, suggestions, and thoughts from the community. GitHub: https://bit.ly/43damDQ Demo: https://bit.ly/3XkKthX https://bit.ly/43damDQ February 24, 2025 at 12:05PM

Show HN: Telescope – an open-source web-based log viewer for logs in ClickHouse https://bit.ly/3QFUd2c

Show HN: Telescope – an open-source web-based log viewer for logs in ClickHouse Hey everyone! I’m working on Telescope - an open-source web-based log viewer designed to make working with logs stored in ClickHouse easier and more intuitive. I wasn’t happy with existing log viewers - most of them force a specific log format, are tied to ingestion pipelines, or are just a small part of a larger platform. Others didn’t display logs the way I wanted. So I decided to build my own lightweight, flexible log viewer - one that actually fits my needs. Check it out: Video demo: https://www.youtube.com/watch?v=5IItMOXwugY GitHub: https://bit.ly/41eJ7pT Live demo: https://bit.ly/41eJ7WV Discord: https://bit.ly/41A4eUU https://bit.ly/41eJ7pT February 26, 2025 at 09:28AM