Show HN for May 14, 2026

In browser PPO training demo, made possible by tinygrad: TinyJit -> WebGPU kernels.

Requires WebGPU.

Every privacy-focused DNS service requires an account: NextDNS, Cloudflare for Families, Apple's iCloud Private Relay (paid, iOS-only). The protocol that doesn’t require one - ODoH - had basically one well-known public relay operator (Frank Denis on Fastly Compute, default in dnscrypt-proxy). I built a second one and the client to talk to it.

An attempt at a single pass LLVM frontend in ~3000 lines of C without external dependencies, malloc, or an AST. Included are some graphical examples. The IR isn't perfect, and the README touches on one particular downfall

Hey HN,

My co-founders and I have been building GridTravel, a free iOS app for planning and sharing travel routes with turn-by-turn GPS nav. We just launched yesterday after App Store approval.

We're three 21-year-old cofounders and best friends since middle school. We built GridTravel after years of frustration navigating new cities on every trip we took together.

The idea: most people either search Google for "top 10 places to visit in…" lists or go on social media to get inspiration on where to go. GridTravel is built around user-generated routes — actual paths someone walked, that you can follow, save, download, and discover from other travelers. Users also have the ability to create private routes and collaborate with their friends.

Tech stack: Mapbox (Nav SDK + maps), Supabase (auth, DB, storage), and Swift. Native iOS for now, Android coming soon.

Our two real cost drivers are Mapbox Search (hit when users create routes) and Mapbox Navigation (hit when users use live navigation). Both have free tiers, then scale with MAU. We launched fully free to remove the barrier to entry. Revisiting pricing in Year 2 once nav costs start burning a hole in our pocket.

Current state: we're in the UGC cold-start hole. The app's value scales with route density in a given city, but route density requires users, who require routes. Classic chicken and egg. Our current plan: 1. Manually seed 25–30 routes per city, starting with 5-10 priority cities where we have personal networks rather than spreading ourselves thin. 2. Short-form content as the primary social channel (TikTok, reels, shorts). Doing A/B testing: whether route walkthroughs convert better than informational/skit videos. 3. Partnering with micro-influencers in those cities (5k-50k following) for in-app routes plus cross-posts on their channels

Curious what HN thinks. Especially anyone who's shipped a UGC product. What worked for you on cold start? What do you wish you'd done differently? Happy to answer any questions about the app, costs, etc.

App link: https://apps.apple.com/us/app/gridtravel-local-routes/id6762...

Robotics is advancing really fast lately, with AI inference, different controllers, software, and parts always changing. I wanted a place that supports many device types, Raspberry Pi, NVDA Jetson, Arduino, ESP32, hardware sources, and maximizes for printability. Instructables, Github, and Thingiverse are currently popular but aren't really focused on robotics, So I built orobot.io to try and make printing robots as standardized and accessible as possible. It uses a lot of Agent built content custom to each project, and every project is designed to be used by humans or your agent.

Features: - Links back to source Github projects - LLMs write descriptions and tips on how to build - View + Download 3d printable STL files in browser - BOM purchase links are kept up to date with LLMs checking Amazon link health - LLMs write Javascript install and controller wrappers custom to each project so a single one-click install works across many frameworks and controller types - Public skill files, clis, and prompts let your agent do everything it needs to walk you through the complexity.

It's still pretty new, so somethings are broken, and there's a lot more I want to build. But I'm very interested to have people try it out let me know if they want to use something like this and give me feedback about where they ran into problems so I can fix it. Thank you HN!

Runo is a web-scraping API that returns typed, structured JSON. You define a schema (field name, type, example value), and Runo fetches the page and returns the data. No HTML, no parsers, no post-processing.

Over the past few weeks, I have been building this non stop. Currently, every scraper API out there solves the site fetching problem but left the extraction of the actual data entirely to users. Runo makes that completely disappear.

For Runo, I went ahead and added JS rendering, stealth mode, and full LLM extraction to make this a fully functional and capable of scraping most if not all sites.

Also, another major problem with current web scrapers is that they charge per feature or bundle them into expensive credit tiers. A single large or JS rendered request can cost 5-75 credits, which means you essentially get nothing out of their plans. Runo is flat per request, no matter the site. At the Scale tier, Runo works out to $0.90 per 1,000 effective requests vs. around $6 for the nearest Firecrawl equivalent. My jaw dropped when I was testing Runo and came across these numbers.

I created a free tier that is 500 requests/month, no credit card required. Take it for a spin and let me what can be improved. I would love feedback.

Permission-first config kit for Claude Code and Google Antigravity IDE. Installs CLAUDE.md + GEMINI.md + 10 specialist agents + 28 skills with one command. MIT license.

A few months ago I stumbled on obra's superpowers repository https://github.com/obra/superpowers. I really liked the approach and idea that you enforce discipline for your agent through a skill-based workflow. Even though coding agents (copilot included) have become a lot better at natively handling complex tasks, they still wander off and lose track of things. I really liked how superpowers fixed this and how it enabled long-running sessions without the agent losing its "focus". So I decided to build a Copilot tailored skill suite around the core idea of superpowers. I didn't just want to port superpowers to Copilot, I took inspiration from it and improved on it. JDS enforces a strict think -> plan -> execute pipeline where nothing gets skipped. It leverages Copilot's built-in sql todo dependencies and provides a live task graph visualizer which helps visualize the agentic workflow and its parallelism. Curious whether others have tried similar approaches, and what's worked or not.

Anthropic announced today that starting June 15, paid Claude plans get a separate monthly credit pool for programmatic usage (claude -p, Agent SDK). Seems OK at first glance but it turns out the monthly credit pool is charged at API rates, which effectively kills any serious programmatic usage for hobbyists.

This is a small Claude Code wrapper which runs claude in a PTY, injecting input, finding the session transcript jsonl file, and using a stop hook to determine when claude is done.

It's a drop-in replacement for claude code. All arguments except the "-p" argument are forwarded as-is.

claude-pee -p "hello world"

Could be used as a cheap way of using Anthropic subscriptions for OpenClaw, Hermes, etc.

Written in Rust, MIT license.

Switched from Claude to ChatGPT? (For whatever reason. Not judging.)

Wish you could get that soothing Claude font and color scheme?

Or maybe you just hate native apps and prefer the bloat of Electron?

Well it's your lucky day, because this is an Electron ChatGPT wrapper that scratches your itch. It even has dark mode. (Applaud here.)

Hi everyone!

This is my first project, so I want to make sure I avoid rookie mistakes in it. Guidelines told me to avoid llm creating this message, so please, accept my apologies for the grammar in advance.

In short, this is an app to unstuck overwhelmed mind and make it get dopamine with something useful instead of reels scrolling.

Choose a category based on your current interest and environment. Choose difficulty level - it is not about completion time, but mostly how uncomfortable for me to do the task. Easy - normal tasks (wash the mug), legendary - I would die to do it without an app (talk to a stranger about zombie apocalypse).

In “not so short”:

The idea popped up from my own struggle: I never been diagnosed with adhd, but my brain is flooded with thoughts and tasks majority of time. Prioritization systems somewhat help but I give up on them and also have a hyper focus superpower that compensates it and I am fine with that. So the issue is not in “Not doing anything” - the real problem is that I am annoyed with the time I extract dopamine from cheap sources like reels. Once my brain gets overwhelmed by the number of thoughts, they collapse into a simple idea of “let’s pour a glass of wine and scroll social media feed for new cool ideas that will change my life magically”.

The idea behind the app is to substitute option of doomscrolling with task completion. I noticed I sincerely enjoy those mental checkmarks so why not to trick the dopamine system into something actually useful? I did that.. hopefully.

This is an early version - very basic, without user profile as I was targeting one singular user - myself. Though, if it helps me, maybe it can help others. With that, next step is to add user profile with filters like marriage (so you don’t have to smile to a cute stranger), pets (or feed the non existing dog), climate (walk 40 mins in Texas summer - wouldn`t recommend), housing (clean garage in your 20th story unit) etc., meanwhile “refresh” button is there for you.

There is a bigger idea behind this, but I’d like to see first if the concept might work for someone else too. And check for possible safety (human and data) concerns before I move forward. User info is stored locally for now, but will be moved to server side once login is introduced.

Also cannot stop thinking, how can I build somewhat social but fare system - no one is checking real life progress, so people will cheat. Doesn’t matter for this concept but definitely blocks social challenges and that’s unfortunate. Will be grateful for some guidance here.

In any way, sorry for lots of words. I hope some of you will find it fun and useful :)

Most AI chat applications (such as ChatGPT or Claude) stream their responses to the client as markdown text. As each new chunk of text arrives, the front end typically re-parses the entire markdown document to render the updated message. This works, but it can quickly slow down the UI for long responses.

I’ve been obsessing over ways to make this more efficient, so I wrote a markdown parser that can parse streaming markdown (semi) incrementally. Instead of re-processing the whole document each time, it only parses what’s new, processing each line only once. Block‑level nodes are buffered until they’re complete (for example, once a paragraph is done and won’t be extended by more text). This also makes parsing the markdown on server possible. The main demo does exactly that. Additionally, animating markdown blocks becomes much simpler and efficient, as a result.

Here’s a demo if you’d like to see it in action: https://markdownparser.vercel.app/experimental

Feel free to type 'Render a table with 10 rows' to see each table row animate in.

I’ve spent a lot of time thinking about this problem, so if you’re working on similar issues, I'd love to chat.

The Shinra Meisin is a modular tracking pipeline for XR systems. It combines eye tracking, mouth/face tracking, SLAM spatial mapping, inside-out body inference, and experimental EEG input into a unified offline-first architecture.

Most XR tracking stacks are fragmented or tied to proprietary hardware, so this project explores a unified, extensible system that gives developers low-level control without relying on closed ecosystems.

Right now, the system has a working eye tracking implementation, with SLAM tracking actively in development. The goal is to make the stack modular enough that it can be adapted or integrated into different headset setups with minimal overhead for developers.

This is being built by a small R&D team at Walker Industries alongside EOZVR, working across perception, tracking, and tooling. I’m currently focused on the Unity/Godot SDK and the companion application layer. The eye tracking system is primarily developed by John, with collaboration across the broader tracking stack. It's been a long road getting things to work, but it's been very rewarding!

We also plan to build a dataset using recorded sessions across a small group of paid volunteers and multiple headset models once the pipeline stabilizes.

Eye tracking demo: https://www.youtube.com/watch?v=QlfCfkzkBB4

I wanted CLI tools like grep to work better with .ipynb files, ie. work on inputs and not on outputs. I created a tools that uses FUSE to strip notebooks on the fly.

The two from parameter golf (one I trained, one was the baseline) are just 16MB each! They produce barely plausible English

It supports: - Image, SF Symbol, and autoplaying video pages - Optional skip behavior - Custom theming - Completion gating - Snapshot-tested SwiftUI UI

I created PandoCast for Windows, for 2 reasons. 1) I was annoyed just enough at intermittent audio hiccups when casting Pandora.com to my soundbar through Chrome tab casting. 2) My professional career has been a marketer and I feel pigeon-holed in a vertical that while I enjoy, it's not what I want to be doing now. I have always been a fiddler, and learned coding and engineering concepts in parallel to my career. I'm looking for a home that values people who have walked different paths, promotes curiosity, and can look at technical problems from entirely new angles.

Wrote this in C#. I came up with an architecture that made sense in my head, did some research and reverse-engineered the new Pandora Modes API, and wrote some basic skeleton code. With the help of Kilo Code, GPT 5.5 (and Gemini) as a programming buddy, I was able to build this, learn about Windows GUI oddities, and fix my annoyance.

Feedback is welcome and appreciated. And if you know of anyone looking for someone like me, I'd love to connect: https://www.linkedin.com/posts/lennychase_github-lennyxcpand...

Open source and self-hosted. Built to keep AI context structured and reusable across MCP compatible clients.

Uses epiplexity selection and meta-rl to find molecules

Hi HN, my first post!

CrunchyCleaner is a minimalist, open-source TUI system cleaner written in Go. I built it because I wanted something faster and cleaner than traditional GUI tools, but more interactive than a simple bash script. Single Binary & Cross-platform (Windows & Linux).