Show HN: Photo.codes – Free, privacy-first photo editor for the web https://bit.ly/4cMmpL7

Show HN: Photo.codes – Free, privacy-first photo editor for the web Hi! I built a free, privacy-first photo editor that runs entirely in your browser. No downloads, no accounts needed. It supports non-destructive editing and lets you create, save, and share custom presets easily. Covers most essential tools like exposure, contrast, color grading, and more. Would love to hear your feedback! https://bit.ly/4cP0RgQ April 26, 2025 at 12:03AM

Show HN: My Python text-based dungeon crawler game engine https://bit.ly/4jwQDEK

Show HN: My Python text-based dungeon crawler game engine I've been recovering from surgery and made use of my downtime by attempting to create a Python dungeon crawler engine. You can find it on GitHub: https://bit.ly/44F4jZp Game worlds are defined in JSON, which provides all the rooms, loot, and monsters to fight. April 25, 2025 at 04:35AM

Show HN: I built Lovable for text bots and mini apps https://bit.ly/4cIJ4Ic

Show HN: I built Lovable for text bots and mini apps Hi HN, During the last weeks, I've been working to create a system that allows you to convert prompts into chatbots and mini apps on platforms that everyone uses on a daily basis. The first integrated platform is Telegram: Telegram is a powerful platform with many integrations and features like bots, apps, games and even payments. So I thought it would be nice to make it easier to create these apps. I created a bot https://bit.ly/4cNqmiA . Workflow is pretty simple: create or choose an existing bot and send a text/voice message about what to create just like you prompt everywhere. To get more instructions, use the /help and /howto commands. Planning to create the same solution for Slack and Discord. Thanks for taking a look. I would love to hear feedback. https://bit.ly/44btuCY x.com/PlutonicDev [1] https://bit.ly/44JuZs7 [2] https://bit.ly/44btva0 https://bit.ly/44btuCY April 24, 2025 at 11:03AM

Show HN: TSCE – Think Before You Speak (Two-Step Contextual Enrichment for LLMs) https://bit.ly/3EqIK4q

Show HN: TSCE – Think Before You Speak (Two-Step Contextual Enrichment for LLMs) Hi HN! I’d like to share TSCE – a Python library that gives small models the ability to follow instructions with above GPT-4-like precision using a two-step architecture. It’s designed for developers who want reliable, high-quality results without the need for fine-tuning large models. TL;DR TSCE uses a two-step approach to improve the reliability of language models. This ensures more accurate and coherent outputs, especially for tasks that require rule adherence or detailed instructions. --- Why You Should Care * No Fine-Tuning: You don't need to fine-tune models. TSCE works with existing models like GPT-3.5 and GPT-4. * Improved Reliability: By decoupling the thinking process (Phase 1) and the answering process (Phase 2), you get more consistent, on-topic results. * Works with OpenAI and Azure: TSCE works out of the box with both OpenAI and Azure OpenAI APIs, so it’s ready to plug into your existing workflows. * Minimal Overhead: The additional latency is small (~2 API calls), but the payoff is huge in terms of clarity and accuracy. Example from tsce import TSCEChat tsce = TSCEChat(model="gpt-3.5-turbo") prompt = "Provide a brief explanation of the benefits of solar energy and how it impacts the environment. Include an example of a renewable energy source other than solar energy." reply = tsce(prompt) print("Answer (Phase 2):", reply.content) # The final response generated after Phase 2 print("Anchor (Phase 1):", reply.anchor) # The structured anchor generated in Phase 1 Output Example: - *Anchor (Phase 1)*: `solar::harvest~light{clean}—echo::sustain[green]` ###END###` - *Answer (Phase 2)*: `"Solar energy offers numerous benefits, particularly in terms of environmental impact. Here are some key points: 1. *Clean Energy Source*: Solar energy is a clean and renewable source of power. It generates electricity without emitting greenhouse gases or pollutants, which helps reduce the overall carbon footprint. 2. *Reduction in Fossil Fuel Dependence*: By harnessing solar energy, we can decrease our reliance on fossil fuels, which are finite and contribute significantly to environmental pollution and climate change. 3. *Sustainable and Renewable*: Solar energy is abundant and inexhaustible. As long as the sun shines, we can harness its power, making it a sustainable energy solution for the long term. 4. *Positive Environmental Impact*: Utilizing solar energy reduces air and water pollution, minimizes habitat destruction, and decreases the strain on natural resources. This contributes to a healthier ecosystem and a more balanced natural environment. *Example of Another Renewable Energy Source*: Wind energy is another excellent example of a renewable energy source. Wind turbines convert the kinetic energy from wind into electrical power. Like solar energy, wind energy is clean, sustainable, and helps reduce greenhouse gas emissions. Wind farms can be established on land or offshore, providing versatile options for generating renewable energy."` --- Installation ``` pip install tsce ``` Customizing System Prompts You can easily customize the system instructions for the final pass by passing in a custom *`final_prefix`*: ``` tsce = TSCEChat(final_prefix="You are a helpful assistant, please respond in 3 sentences.") reply = tsce("Explain the concept of recursion.") print(reply.content) ``` --- If you have any questions or want to try it out, feel free to comment here or head over to the repo. [GitHub Repo]( https://bit.ly/4jqV9UY ) [GDrive: Read the paper, See the proof]( https://bit.ly/3EDZZPG ) Looking forward to hearing what you think! https://bit.ly/4jqV9UY April 25, 2025 at 01:13AM

Show HN: Faasta – A self-hosted Serverless platform for WASM-wasi-HTTP in Rust https://bit.ly/3YgEFX5

Show HN: Faasta – A self-hosted Serverless platform for WASM-wasi-HTTP in Rust I've just released an early version off my project I've been working on for a few months now and would love some feedback. https://bit.ly/44EYpHJ I was surprised there isn't yet an open/source and standards compliant way to host wasi-http functions in a way that takes advantages of WASM, a multi tenanted application. If you're not familiar with wasi, Compared to something like AWS Lambda, this approach is much more efficient as a single process can serve 1000s of function invocations concurrently and asynchronously, instead requiring an entire VM. This is still early days for the project, but feel free to download the cli utility with cargo install cargo-faasta. Feel free to test deploying functions on my hosted instance at https://bit.ly/4cTNWuq . The service is free to use and currently supports deployments via GitHub OAuth, with a limit of 10 functions per GitHub account. https://bit.ly/44EYpHJ April 25, 2025 at 01:31AM

Show HN: My from-scratch OS kernel that runs DOOM https://bit.ly/42oTBoC

Show HN: My from-scratch OS kernel that runs DOOM https://bit.ly/4jPmtfC April 24, 2025 at 01:15AM

Show HN: An all-in-one blog for learning Large Language Models (LLMs) https://bit.ly/4jnpXpE

Show HN: An all-in-one blog for learning Large Language Models (LLMs) An all-in-one blog for learning LLM ins and outs: tokenize, attention, PE, and more Project I've been diving deep into the internals of Large Language Models (LLMs) and started documenting my findings. My blog covers topics like: Tokenization techniques (e.g., BBPE) Attention mechanism (e.g. MHA, MQA, MLA) Positional encoding and extrapolation (e.g. RoPE, NTK-aware interpolation, YaRN) Architecture details of models like QWen, LLaMA Training methods including SFT and Reinforcement Learning If you're interested in the nuts and bolts of LLMs, feel free to check it out: https://bit.ly/4jrd7qy https://bit.ly/4jQGhPQ April 23, 2025 at 11:21PM

Show HN: Talk to 3000 ICLR 2025 (machine learning conference) papers https://bit.ly/4isj98U

Show HN: Talk to 3000 ICLR 2025 (machine learning conference) papers ICLR is one of the top machine learning / AI conferences. We put the over 3000+ accepted papers into a “Pod” which you can chat. Example queries: “find papers with more than one OpenAI-affiliated author” “find papers that propose alternatives to Transformer architecture in LLM” “give an overview of all spotlight or oral papers with Yann Lecun as author” RadPod is an agentic AI system that supports reasoning over a large amount of context, in this case over 90M tokens, with high accuracy. https://bit.ly/4cJhyKT April 23, 2025 at 02:21AM

Show HN: Trailmarks – Huge, browser-based, Carmen Sandiego-style travel game https://bit.ly/42oxSNv

Show HN: Trailmarks – Huge, browser-based, Carmen Sandiego-style travel game I'd love to have people test-play my Carmen Sandiego/Backpacker style travel game. It's a retro genre... but if you like geography and quizzes, you'll like this! It's a working alpha with most of the content I intend to add, but I need to have people give it a try now before I decide which directions to fully develop! I'm torn between putting effort into missions, more advanced in-game trading economy or creating more of a narrative. But I'm also curious if you get the navigation, if the core gameplay makes sense and if you want to continue playing, basically. Hope you find it worth playing. It's a pure browser game with no login, you're playing for free immediately in this alpha! Please feel welcome to DM feedback or reply or anything! https://bit.ly/4lJ2ZLd April 23, 2025 at 12:55AM

Show HN: Durable Python Workflows https://bit.ly/3EJX73E

Show HN: Durable Python Workflows https://bit.ly/443gXRE April 22, 2025 at 11:41PM

Show HN: Onmom – AI meal planner and calorie tracker that works from text/photo https://bit.ly/4ixt0dL

Show HN: Onmom – AI meal planner and calorie tracker that works from text/photo I built OnMom because I was tired of apps like MyFitnessPal being too tedious for meal logging. This one uses AI to analyze meals from quick text or photo input and gives suggestions. It’s free and still super early – would love feedback on what’s missing or confusing! https://bit.ly/3GmJmIY April 22, 2025 at 12:56AM

Show HN: I made TypeScript's type inference more strict (and smarter) https://bit.ly/4it6xP6

Show HN: I made TypeScript's type inference more strict (and smarter) As a TypeScript developer, I often found myself wishing the type system could do more—*especially when omitting or modifying deeply nested properties* inside complex objects and arrays. For instance, what if I want to remove a deeply nested field like `user.profile.email` and also something like `user.posts[ ].meta.shares` from a type? TypeScript doesn't really provide a built-in way to do that. So I built *DeepStrictTypes* — a utility that lets you *omit deeply nested keys*, even inside arrays, with full type inference and strictness. Here’s an example: ```ts type Example = { user: { id: string; profile: { name: string; age: number; email: string; }; posts: { title: string; content: string; meta: { likes: number; shares: number; }; }[]; }; }; // Remove 'user.profile.email' and 'user.posts[ ].meta.shares' type Omitted = DeepStrictOmit< Example, 'user.profile.email' | 'user.posts[*].meta.shares' >; ``` The resulting type: ```ts { user: { id: string; profile: { name: string; age: number; }; posts: { title: string; content: string; meta: { likes: number; }; }[]; }; } ``` Works great for: - Cleaning up types for API responses - Dynamically transforming deeply nested data - Improving type safety when handling structured JSON [ https://bit.ly/3S6gjM1 ]( https://bit.ly/3S6gjM1 ) Would love your feedback or ideas for improvements! https://bit.ly/3GyfgSz April 22, 2025 at 04:47AM

Show HN: Prompt Coded 3D Asteroids https://bit.ly/42GNg6I

Show HN: Prompt Coded 3D Asteroids https://bit.ly/42UUohc April 21, 2025 at 11:55PM

Show HN: I Made the Duolingo for Wine – Wine Bible https://bit.ly/4iw8dY1

Show HN: I Made the Duolingo for Wine – Wine Bible I recently launched Wine Bible, a mobile app for learning about wine through interactive lessons, quizzes and tasting guides. I built it because most wine education is either buried in books or locked behind expensive courses. I wanted a way to help people learn wine the same way they’d learn a language — through daily practice, fun repetition, and interactive lessons. Right now, Wine Bible includes: • A growing library of wine lessons (grapes, regions, wine making, wine business, sparkling wine, fortified wine) • Guided wine tasting with an AI-powered coach • Over 250 grape varieties with their taste notes, origin, history, food pairing and regions Would love feedback from anyone curious about wine, building in edtech, or who’s worked on similar educational apps. https://bit.ly/4jgAaEd April 21, 2025 at 11:58AM

Show HN: Comparelists.org – Instantly Compare Two Lists, Find Differences https://bit.ly/3Y4ZiWb

Show HN: Comparelists.org – Instantly Compare Two Lists, Find Differences Hey HN, I got tired of manually comparing two lists (think: emails, product SKUs, code, whatever) and built CompareLists.org to make it painless. You just paste your two lists, hit compare, and instantly see what’s unique to each list, what matches, and any duplicates. It handles thousands of lines, works right in your browser (no data leaves your device), and you can export results as CSV/TXT/JSON. There are options for case sensitivity, whitespace, and more. It’s free and there’s no signup. I’d love for you to try it out and let me know what you think—or if there’s a feature you wish it had. Feedback (and bug reports) super welcome! Thanks! https://bit.ly/4jmbSJc April 21, 2025 at 01:25AM

Show HN: Keep your PyTorch model in VRAM by hot swapping code https://bit.ly/3GuYVxX

Show HN: Keep your PyTorch model in VRAM by hot swapping code https://bit.ly/3Y7ANaO April 21, 2025 at 01:21AM

Show HN: LLM Shell Tools – AI-powered command line helpers(open source + local) https://bit.ly/4435FwW

Show HN: LLM Shell Tools – AI-powered command line helpers(open source + local) Hey HN, I ve been playing around with LLM integrations for the shell and I have a tiny set of shell tools that use LLMs to improve my CLI experience. I’ve been using it while building and it’s already proving usefu to me. Right now it includes: 1. Command Not Found Hook Catches “command not found” errors and asks an LLM to suggest what might have gone wrong (e.g. typos, missing tools, etc), works for bash and zsh 2. Git Commit LLM A wrapper for git commit that suggests clearer commit messages based on your staged changes. It’s helped me avoid a bunch of lazy messages like “fix stuff”. Prereqs: • OpenAI API key • Bash or Zsh • curl and jq Still minimal and hacky—but I’d love feedback or ideas. Especially curious if others would find this helpful or have feature suggestions. Repo: https://bit.ly/42P6Yyt https://bit.ly/42P6Yyt April 20, 2025 at 09:04AM

Show HN: Hyprnote – VSCode for Meeting Notes (Open-Source and Local-First) https://bit.ly/440Wjlt

Show HN: Hyprnote – VSCode for Meeting Notes (Open-Source and Local-First) Hi HN. This is Yujong from Hyprnote. Hyprnote is an open-source, local-first macOS app that you can use for meetings. - It uses Mic + System Audio for audio sources. - Whisper (small-en-q8) + Llama (3b-q8) for transcription + summarization. Some interesting bits: - Extension & plugin system that you can use to create widgets (inspired by VSCode). - Built with Libsql. Will enable DB sync across cloud/devices in the future. There are still some rough edges, but I thought it is worth sharing the concept and getting feedback. We have an interesting onboarding video. Please try it out! ```bash brew tap fastrepl/hyprnote && brew install hyprnote --cask ``` If you want to see how it looks first - https://www.youtube.com/watch?v=-apfueHQHBk https://bit.ly/4jfgS25 April 20, 2025 at 05:39AM

Show HN: EyesOff – Alerts you when someone peeps at your screen https://bit.ly/42pXkSY

Show HN: EyesOff – Alerts you when someone peeps at your screen Hey HN, this is Yusuf. I've built a privacy focused macOS app which makes use of a locally running neural network (YuNet), to notify you if other people are looking at your screen. YuNet runs fully on-device with no data leaving your computer. The app utilises a 230kb facial detection model, which takes images from your webcam and checks for any faces entering the viewing field of your webcam. If the number of faces exceeds the threshold an alert will be shown. Built with Python + PyQt, the YuNet code comes from OpenCV. Currently it's a macOS app only, however I will be widening access to windows devices soon. Link + Source code: https://bit.ly/4cLFD3O I'd love your feedback on the app, I look forward to reading your comments on thoughts and future directions you'd like to see! https://bit.ly/4cLFD3O April 19, 2025 at 08:44PM

Show HN: Make your bookmarks smarter with AI https://bit.ly/4ihLn6d

Show HN: Make your bookmarks smarter with AI https://apple.co/42c6CBP April 19, 2025 at 04:55AM