Show HN for November 10, 2025

Hi HN,

I’m Rajiv, a software engineer turned Math teacher living in the mountains, where I like to slow down life while still building useful software.

I recently built DroidDock, a lightweight and modern macOS desktop app that lets you browse and manage files on your Android device via ADB. After 12 years in software development, I wanted a free, clean, and efficient tool because existing solutions were either paid, clunky, or bloated.

Features include multiple view modes, thumbnail previews for images/videos, intuitive file search, file upload/download, and keyboard shortcuts. The backend uses Rust and Tauri for performance.

You can download the latest .dmg from the landing page here: https://rajivm1991.github.io/DroidDock/ Source code is available on GitHub: https://github.com/rajivm1991/DroidDock

I’d appreciate your feedback on usability, missing features, or bugs. Thanks for checking it out!

— Rajiv

Hi HN,

We’re Ruben, Afnan, and Theo, and we’re building Davia to solve a common problem: documenting and explaining large codebases is complex. It takes too long to generate even a first draft of a wiki, visuals are essential to understand the structure, internal docs should be editable in the IDE, and most solutions aren’t open.

Davia is an open source tool. You enter the path of your repo, and it generates a visual wiki you can explore and edit. Diagrams are created automatically, and you can update everything either in your IDE or in a Notion-like editor.

The project is still early, and we’d love to hear feedback, ideas, or experiences from anyone interested in documenting and sharing code internally.

GitHub: https://github.com/davialabs/davia

Hi HN! I built LLM OneStop (https://www.llmonestop.com), a unified interface for accessing multiple AI language models in one place. The main problem I wanted to solve: constantly switching between different AI platforms, managing multiple subscriptions, and losing conversation context when comparing outputs across models. Key features:

Switch between GPT-4, Claude, Gemini, Llama, and other models mid-conversation Compare responses side-by-side Single interface instead of juggling multiple tabs/subscriptions Free tier available to try it out (no credit card needed) "Connect" plan lets you bring your own API keys for unlimited usage

I built this because I found myself constantly copy-pasting prompts between ChatGPT, Claude, and other platforms when I wanted to compare responses or find the best model for a specific task. You can try it right now with the free tier - would love to hear your feedback on what works, what doesn't, and what features you'd like to see!

Just shipped a tiny “eyes follow your cursor” widget with a live playground so you can noodle around. The hosted embed has a paid tier for production apps, but the demo is open—looking for ideas on fun interactions or motion tweaks.

Git AI is a side project I created to track AI-generated code in our repos from development, through PRs, and into production. It does not just count lines, it keeps track of them as your code evolves, gets refactored and the git history gets rewritten.

Think 'git blame' but for AI code. There's a lot about how it works in the post, but wanted to share how it's been impacting me + my team:

- I find I review AI code very differently than human code. Being able to see the prompts my colleagues used, what the AI wrote, and where they stepped in to override has been extraordinarily helpful. This is still very manual today, but hope to build more UI around it soon.

- “Why is this here?” — more than once I’ve giving my coding agent access to the past prompts that generated code I’m looking at, which lets the Agent know what my colleague was thinking when they made the change. Engineers talk to AI all day now…their prompts are sort of like a log of thoughts :)

- I pay a lot of attention to the lines generated for every 1 accepted ratio. If it gets up over 4 or 5 it means I’m well outside the AI’s distribution or prompting poorly — either way, it’s a good cause for reflection and I’ve learned a lot about collaborating with LLMs.

This has been really fun to build, especially because some amazing contributors who were working on similar projects came together and directed their efforts towards Git AI shine. We hope you like it.

This was a goofy little project I worked on. It's fully open source and you can check it out via the Github link.

Hi HN, I built DeepFakeCheck.com a lightweight web tool that lets you upload an image and immediately check if it’s AI-generated or real, powered by the SightEngine API under the hood. Why I made it: Deepfakes are increasingly used (misinfo, identity fraud) and most people have no quick way to verify images. I wanted something simple, no account needed initially, and transparent. For this test launch I will manually increase the free scans to 25 (instead of the usual 3) after sign up so you can test it thoroughly for free and give feedback. Please be patient while I check the signups as much as possible and manually increase the credits from 3 to 25 (usually within minutes) unless I am sleeping. Try it here: DeepFakeCheck.com What I’d love: Your feedback on accuracy, UI, friction, what you’d fix. Any ideas for integrations (Slack-bot, API, browser extension) or real-world use cases. Any issues, edge cases you’ve found. Technical notes: Built with Next.js + Tailwind and deployed on Vercel. SightEngine API for detection. Redis via Upstash for rate limit / caching of free scans. No images are stored: all uploads only live briefly for processing, then discarded. Thanks for checking it out. I’ll be actively monitoring comments here and open to follow-up improvements. Rody

I know, I know: another resume generator.

But this one is simple, easy to customize, and gets to the point: a nicely formatted HTML or PDF document you can share.

Note: Linux and MacOS friendly; not tested on Windows.

The giTshirt story A little while ago, I decided I needed a way to commemorate my first SaaS. Sure, I had a 3D printed plate of its name, but that was not enough, I needed something more…

While that thought lingered at the back of my mind, I saw a funny tweet showcasing ridiculous commit messages. I thought, “Hey! I have some stupid commit messages, let me go screenshot them and share them as a reply to that tweet.” Somehow the two thoughts connected. It instantly clicked: What if I could wear my stupid git commit messages? What if I put it on a T-shirt? A giTshirt…

That would be the perfect way to showcase what I’ve made, as well as a funny conversation starter at conventions. So I developed an algorithm that places commit messages from repositories you’ve chosen and creates a ready-to-order T-shirt. Then I worked with a designer to create a logo and a special card that you receive with your giTshirt. I didn't stop reworking giTshirt until I created something I'D like to wear.

Spoiler - It took me weeks but I managed to do it.

The giTshirt you see on the home page is the one I ordered and the very one I'm wearing while typing. (It's pretty cool!)

How it all works Here's a step-by-step of everything that happens from clicking “Generate giTshirt” to getting it to you: 1. Security: 1.1 Stop you if you are not authenticated 1.2 Stop you if you are generating 3 shirts per minute (pls don't :)) 1.3 (Sadly) Stop you if there are more than 300 commits selected. You just physically can't fit more than 300 messages on a T-shirt, sorry 2. Generating images: 2.1 I needed to create a 2D collision system to place the commit messages 2.2 The commit messages are placed at almost random 2.3 There are 4 total images generated. Back, front, left sleeve, and right sleeve 3. Creating the product 3.1 I upload the 4 images to Cloudinary 3.2 These images are then downloaded to my T-shirt provider 3.3 The code places them on the T-shirts 3.4 Then a product is created and I get back the mockups back 3.5 These mockups are then displayed to the user before a purchase is made

Here's how you can get a giTshirt in less than a minute: This is not a marketing gimmick; I timed it, and it took me 43 seconds. 1. Sign up with GitHub 2. Select a repository - both public and private ones work 3. (optional) All commit messages are pre-selected, but you could nit-pick the ones you want 4. Click on the "Generate giTshirt" button 5. Order and wait!

If you like the idea and want to order one you can visit - https://giTshirt.com If you like the idea but don't want to order one you can always visit the ProductHunt launch and support giTshirt there - https://www.producthunt.com/products/gitshirt Thanks for reading!

I built UniWorld V2, a next-generation AI image editing model that understands regions, text, and context — all in one coherent workflow.

Unlike typical diffusion editors, UniWorld V2 applies precise regional edits, integrates reinforcement-learning feedback (Edit-R1), and treats text as a native visual element — not just texture.

Key Features

• Region-Aware Editing – Mask any area, apply a prompt; lighting and global coherence stay intact.

• RL-Enhanced Accuracy (UniWorld-R1) – MLLM-based reward model improves intent alignment and edit quality (outperforms GPT-Image-1, Nano Banana, Gemini).

• Multi-Round Edit Consistency – Edit → re-edit → refine without style drift.

• Advanced Typography Editing – Insert or replace text while preserving font, spacing, and perspective.

• Precision Object Control – Move, add, remove, or replace objects with explicit commands.

Use Cases

Ad & social asset localization

Product & UI iteration

Education / L&D content

Editorial & newsroom visuals

E-commerce & creator workflows

https://www.uniworldv2.com/?i=d1d5k

UniWorld V2 combines region-aware control, RL precision, and advanced typography — setting a new benchmark for AI-powered image editing tools.

Would love feedback from the HN community — especially around usability, edit stability, and RL feedback design.

Hi HN,

I've recently published my first iOS app to the Appstore! Its a photo cleanup app which uses on-device Apple Vision models to cluster similar images together into groups, and then ranks each photo within each group.

Link: https://apps.apple.com/app/ai-photo-cleaner-sorter/id6754172...

How it works: The best photo from each group gets marked as "Saved" by default, however in the Review stage, you can select others from the group to save if you want or unselect the "Best" one and choose another one to keep. The general concept is "keep the best, remove the rest", and in theory if you agree with the clustering/scoring and didn't want to make any amendments, you could clean up 100's of photos (and MBs) with just a few taps.

Motivation: I'm one of those people who somehow builds up an insanely bloated camera roll - 15 photos of the same restaurant dish, 10 versions of the same sunset. So I initially built this for myself to help cleanup those 500 photos taken during a 1 week vacation... I've tried a load of the other apps out there, but most seem to have adopted a Tinder style process - swipe left, swipe right, which hardly decreases the workload. I wanted an opinionated way to do a mass cleanup easily.

Things on my roadmap: - Localisation for major languages - Improved background processing (so you can clean up 1,000+ photos without keeping the app open) - User preferences for how the photo scoring is calculated i.e. some people value emotional signals over image sharpness.

Its totally free to use (maybe one day I'll try and monetise it, if I'm ever fully satisfied )), I'd love any feedback - good and bad! (Especially would appreciate feedback on how well the grouping and scoring performs for you - any quirks?)

Thanks! Nick

Quick grammar, punctuation, and spelling checker. Free, no ads, no sign-up. Paste up to 1000 words, get instant suggestions with explanations. Built for writers who need fast, accurate checks without friction. One-click corrections with grammar checker results. Feedback welcome—what features would make it more useful?

AI Agents in GoLang, open-source, cloud-native, GitOps-enabled, everything you need to start fast, run it where you want: local, your cluster or ours.

React Chrono v3.0 is a major release for the popular React timeline component.

Highlights: - Migrated to Vanilla Extract for zero-runtime CSS (no styled-components) - New grouped configuration API (props: layout, interaction, content, display, media, animation, accessibility, i18n) - Per-element Google Fonts control and font caching - Complete i18n covering 40+ text elements - Cross-browser fullscreen mode with keyboard shortcuts - Advanced dark mode (36 customizable theme props, 13 new in v3.0) - Sticky toolbar, enhanced search - TypeScript, React 18.2+ support - Full backward compatibility with v2.x

More details in the README and changelog!

I quit my well-paid job in a good company to work on my “next big thing” startup, with ~6 months of runway and the usual plan: ship in a month, then do marketing and SEO. Reality: the launch kept slipping, my runway kept shrinking, and the SEO tools I needed started at ~$200/month subs — which, at that point, was basically “one extra month of life” money. So I asked my wife (also a dev) to vibecode a minimal internal rank tracker for me. Two weeks later it worked well, so in 2 extra weeks we shipped a production-grade product. Looking for feedback.

A few months ago I realized I had no idea where our money was actually going. Not in a "we're broke" way, just... groceries, random Amazon purchases, subscriptions I forgot about, it all blurred together.

I tried a few existing apps but they were either too complex (why does budgeting software feel like filing taxes?) or wanted to sell me credit cards or were expensive and had subscription. Nothing answered the simple question: "who in our household is spending what, and where?"

The main insight: my partner and I just wanted to see patterns without judgment. Like "oh, I spent $200 on coffee this month" or "wait, we bought HOW many toys?" or "this grocery store is bleeding us dry, let's switch."

So I built something dead simple: enter expense, amount, who it's for, merchant, category. Done. Takes 10-20 seconds. The manual entry actually makes you more aware of spending, and then you get nice visual breakdowns.

No guilt trips, no "you're $5 over budget!" notifications, no gamification. Just clean data.

Plot twist: we now actually enjoy looking at our spending together. Turns out money conversations are way easier when you're not guessing or arguing over who bought what by guessing the number!

It's free to use. Would genuinely love feedback. What works, what doesn't, what's missing. I'm actively building and happy to add features that actually make sense.

See: calmspend.com

This pattern principle is almost unknown. In the program, press Q and then select the Chinese character to learn more about it.