Hacker News

Article URL: https://emojicombo.net Comments URL: https://news.ycombinator.com/item?id=44950772 Points: 1 # Comments: 0

Honest question: Do you also give AI minimal info and expect magic?We’ve all been there: Ask AI to code something. Get code that works but looks nothing like your project. Spend hours fixing it to match your style.I built this tool for Claude Code to fix that problem. It’s like having AI that actually reads your codebase first.How it works:1. Tell it what you want in plain English 2. It asks questions if it needs more info 3. It studies your existing code patterns 4. It searches the web for best practices if needed 5. It creates a detailed plan for Claude Code to followResult: Code that matches your project from day one.Claude Code gets a complete guide instead of a vague prompt. It knows your naming style, your architecture, your testing setup. Everything.Works with any programming language. The tool figures out your project automatically.No more random AI code. No more refactoring. Just code that fits.Built for Claude Code CLI but the approach works anywhere. Would love to hear what you think. Comments URL: https://news.ycombinator.com/item?id=44950769 Points: 1 # Comments: 0

Article URL: https://github.com/Araq/ormin Comments URL: https://news.ycombinator.com/item?id=44950753 Points: 1 # Comments: 0

Article URL: https://victorwynne.com/vision-for-ipad/ Comments URL: https://news.ycombinator.com/item?id=44950742 Points: 2 # Comments: 0

Within the single context of writing code, we’ve had quite a few attempts throught history to generate code. Often times within very rigid constraints.GenAI is fits in the same space, but with extra steps, benefits and drawbacks. It’s not “a junior dev”. It’s a new hammer.The craftsman enjoys the new hammer, pushes its limits, nerds out about the intricacies. Tools have limits in terms of wear and tear, plus cost.Junior devs are humans, looking to survive and flourish. They pick up new tools faster than most. The only barrier the same that has always been: access to said tools and visibility over the outcome.What’s the benefit of over-anthropomorphizing a hammer? Comments URL: https://news.ycombinator.com/item?id=44950732 Points: 2 # Comments: 4

Hey HN, I’m Brandon, developer for Harbor Guard, an open-source container security platform that unifies multiple scanning tools into a single intuitive dashboard.Over the past month, I got tired of juggling different CLI scanners (Trivy, Grype, Syft, Dockle, etc.), for my work in DevSecOps. Stitching results together started becoming a hassle and my organization started heavily focusing on container security. It was messy, repetitive, and hard to keep track of active vulnerabilities.So I built Harbor Guard, a modern, Next.js app that runs all major open-source container scanners, stores the results, and turns them into interactive visualizations you can actually use.What Harbor Guard does:- Runs 6 major tools: Trivy, Grype, Syft, Dockle, OSV Scanner, and Dive- Unified dashboard: View and compare results without jumping between CLIs- Historical tracking: Persist and compare scans across versions- Advanced visualization: Interactive scatterplots, severity-based grouping, and layer-by-layer analysis- Report export: Grab JSON or full ZIP packages for compliance- API access: REST endpoints for automation and integrations- Docker-native: Run it with a single docker run -p 3000:3000 …Personally, I needed a tool that made vulnerability data easier to explore visually and had some fun putting something together.Try it outGitHub: https://github.com/HarborGuard/HarborGuardLive Demo: https://demo.harborguard.coIt’s free, open source, and I’d love feedback, bug reports, or feature ideas. Contributions welcome too.Thanks, Brandon Comments URL: https://news.ycombinator.com/item?id=44950704 Points: 1 # Comments: 0

Hi HN,I’m excited to launch AICMaker (AI Original Character Maker), a tool we built to solve a persistent problem in AI image generation: character consistency.If you’ve ever tried to create a story or game concept with an AI, you know the frustration. You generate a great character, but you can’t get the same character from different angles, in different outfits, or with different expressions. You end up with a collection of similar-looking strangers, not a consistent persona.AICMaker fixes this. Our tool lets you create a “character core” and then generate that exact same original character in any pose, style, or scene while maintaining their core identity (face, body type, etc.). It’s designed for indie game developers, writers, animators, and TTRPG enthusiasts who need reliable visual assets.How it works (the tech stack):We fine-tune a core model on specific techniques for identity preservation.The process involves creating a custom LoRA (Low-Rank Adaptation) or a similar concept for your character, but wrapped in a simple UI—no need to understand the underlying ML concepts.You start with a few base images (or generate one from a text prompt), and the system learns the character’s identity to ensure all future renders are consistent.Why we built it: As game developers ourselves, we found existing tools either too technically complex (requiring local SD setup and manual LoRA training) or too simplistic (failing at true consistency). We wanted a web-based, accessible tool that just works.Try it out / Examples: You can create a character and generate a few images for free.Link: https://aiocmaker.com/Example Gallery: [Link to an example gallery if you have one]We’d love your feedback. This is very much an MVP, and the insights from this community are invaluable. What are your biggest pain points with generative AI for character design? Does our approach make sense?Looking forward to your thoughts and questions. Comments URL: https://news.ycombinator.com/item?id=44950703 Points: 1 # Comments: 0

Article URL: https://arxiv.org/abs/2507.16018 Comments URL: https://news.ycombinator.com/item?id=44950702 Points: 1 # Comments: 0

Article URL: https://xquant.substack.com/p/an-introduction-to-the-market-making Comments URL: https://news.ycombinator.com/item?id=44950700 Points: 1 # Comments: 0

Article URL: https://torrentfreak.com/ad-blocking-is-not-piracy-decision-overturned-by-top-german-court-250819/ Comments URL: https://news.ycombinator.com/item?id=44950695 Points: 3 # Comments: 0

Article URL: https://medium.com/@bulanramai2558/chinese-room-experiment-vs-supat-language-room-experiment-experiment-1-claude-acc8022ef9df Comments URL: https://news.ycombinator.com/item?id=44950682 Points: 1 # Comments: 0

Article URL: https://www.theregister.com/2025/08/19/palo_alto_networks_q4_fy25/ Comments URL: https://news.ycombinator.com/item?id=44950676 Points: 2 # Comments: 0

Article URL: https://andymasley.substack.com/p/a-ton-of-ai-images-ive-made-that Comments URL: https://news.ycombinator.com/item?id=44950671 Points: 1 # Comments: 0

Hi HN, we are Zaid, Muhammad and Hammad, the co-founders of Uplift AI (https://upliftai.org). We build models that speak underserved languages — today: Urdu, Sindhi, and Balochi.A billion people worldwide can’t read. In countries like Pakistan – the 5th most populous country – 42% of adults are illiterate. This holds back the entire economy: patients can’t read medical reports, parents can’t help with homework, banks can’t go fully digital, farmers can’t research best practices, and people memorize smartphone app button sequences. Voice AI interfaces can fix all of this, and we think this will perhaps be one of the great benefits of modern AI.Right now, existing voice models barely work for these languages, and big tech is moving slowly.Uplift AI was originally a side project to make datasets for translation and voice models. For us it was a “cool side-thing” to work on, not an “important full-time thing” to work on. With some initial data we hacked together a Urdu Voice Bot on Whatsapp and gave it to one domestic worker. In two days 800 people were using it. When we dived deeper into understanding the users, we learned that text interfaces don’t work for sooo many. So we started Uplift AI to solve this problem fulltime.The most challenging part is that all the building blocks needed for great voice models are broken for these languages. For example, if you are creating a speech synthesis model, you will scrape a lot of data from youtube and auto-label it using a transcription model… all very easy to do in English. But it doesn’t work in under-served languages because the transcription modes are not accurate.There are many other challenges. Like when you hire human transcribers to label the data, often they don’t have any spell correctors for their languages, and this creates lots of noise in the data… making it hard to train models with low data. There are many more challenges in phonemes, silence detection, diacritization etc.We solve these problems by making great internal tooling to help with data labeling. Also, we source our own data and don’t buy it. This is counterintuitive, but a big advantage over companies buying data and then training. By sourcing our own data we create the right data distributions and get much better models with much less data. By doing the entire thing inhouse, (data, labeling, training, deploying) we are able to make a lot faster progress.Today we publicly offer a text to speech APIs for Urdu, Sindhi, and Balochi. Here’s a video which shows this: https://www.loom.com/share/dcd5020967444c228e9c127151e7a9f5.Khan Academy is using our tech to dub videos to Urdu (https://ur.khanacademy.org).Our models excel at informational use cases (like AI bots) but need more work in emotive use-cases like poetry.We have been giving a lot of people private access in beta mode, and today are launching our models publicly. We believe this will be the fastest way for us to learn about areas that are not performing well so we can fix them quickly.We’d love to hear from all of you, especially around your experiences with under-served languages (not just the Pakistani ones we’re starting with) and your comments in general. Comments URL: https://news.ycombinator.com/item?id=44950661 Points: 1 # Comments: 0

Article URL: https://github.com/DRYTRIX/RequirementsTracker Comments URL: https://news.ycombinator.com/item?id=44950635 Points: 1 # Comments: 0

Article URL: https://machinelearningmastery.com/making-sense-of-text-with-decision-trees/ Comments URL: https://news.ycombinator.com/item?id=44950633 Points: 1 # Comments: 0

Article URL: https://english.elpais.com/lifestyle/2025-08-17/the-steps-we-take-the-hours-we-sleep-the-books-we-read-why-are-we-counting-everything.html Comments URL: https://news.ycombinator.com/item?id=44950631 Points: 3 # Comments: 0

Article URL: https://www.theverge.com/news/761240/uk-apple-us-encryption-back-door-demands-dropped Comments URL: https://news.ycombinator.com/item?id=44950600 Points: 4 # Comments: 0

Article URL: https://getparthenon.com/blog/ai-is-a-junior-dev-and-needs-a-lead/ Comments URL: https://news.ycombinator.com/item?id=44950599 Points: 3 # Comments: 0

Article URL: https://www.telegraph.co.uk/business/2025/08/19/uk-caves-in-to-trump-demands-on-apple-back-door/ Comments URL: https://news.ycombinator.com/item?id=44950591 Points: 1 # Comments: 0