15:[[48134269,{"by":"levmiseri","descendants":0,"id":48134269,"kids":[48134351],"score":1,"text":"This is more of an experiment/demo than something serious, but wanted to share and see what people think.

The main premise is that you can create anything by typing "/create {anything}" and the game will do its best to produce such object with all the functionality and properties that would make sense.

It might look simple at first glance, but the underlying systems are fairly complex. Create dragons you can fly on, spells, weapons, factories, ...","time":1778760600,"title":"Show HN: Asciidia – LLM Crafting Game","type":"story","url":"https://asciidia.com"}],[48051142,{"by":"Yizahi","descendants":1,"id":48051142,"kids":[48051143,48051182],"score":3,"time":1778170195,"title":"Claude Code creator Boris Cherny is sick of the phrase 'vibe coding'","type":"story","url":"https://www.businessinsider.com/claude-code-creator-boris-cherny-vibe-coding-anthropic-ai-2026-5"}],[48006743,{"by":"aminekhd","descendants":0,"id":48006743,"kids":[48006966],"score":2,"text":"I own a code editor at https://pterocos.eu.org - it's an online HTML/CSS/JS editor with:

No accounts required

No tracking or analytics

All user data stored in browser localStorage (zero server-side storage)

Looking for feedback on the editor experience. Does the privacy focus matter to other developers here?","time":1777889900,"title":"Ask HN: Feedback on my privacy-focused code editor?","type":"story"}],[47988241,{"by":"usermac","descendants":1,"id":47988241,"kids":[47988538,47988258,47988406],"score":2,"text":"I deal with the professional LLMs, of course, but I'm really intrigued by the possibility of local coding offline. I've got a MacBook Air M4 16gb. Does it have any chance at all of doing coding? NOTE: I am not too worried about the context window because the way I work is very targeted and surgical. I'll have it look at one file and have it do something very exact to that file.","time":1777741712,"title":"Ask HN: Are small local LLMs good at coding?","type":"story"}],[47983011,{"by":"brady_bastian","descendants":0,"id":47983011,"kids":[47983065],"score":2,"text":"Parsing FHIR data and converting that to OMOP standard is a notoriously difficult problem. I think we've cracked it.

I've open sourced two GitHub repos:

1. Forge-Core. This is the secret sauce. Automatically parse ANY json structure inside your own data warehouse by mapping your schema and building dbt models on the fly. Easy to set up and use. https://github.com/foxtrotcommunications/foxtrotcommunicatio...

2. Avalon-Public. Following the forge parse, use these dbt models to correctly and efficiently build your needed omop models. https://github.com/foxtrotcommunications/foxtrotcommunicatio...","time":1777692234,"title":"Show HN: Open-Source FHIR –> OMOP Pipeline","type":"story","url":"https://forge.foxtrotcommunications.net/"}],[47977625,{"by":"flamestro","descendants":0,"id":47977625,"kids":[47977682],"score":1,"text":"Sanishne is a lightweight tool for teams to save, tag, and find shared links without losing them in Slack/Discord. It uses simple shared boards with viewer/editor/owner roles, and focuses only on bookmarks (no docs, no bloat). Free and currently an early MVP.

Would love to get some feedback.","time":1777657072,"title":"Show HN: Sanishne – shared bookmarks for teams (so links don't vanish in Slack)","type":"story","url":"https://sanishne.org/"}],[47972888,{"by":"mayerwin","descendants":40,"id":47972888,"kids":[47973783,47973460,47976910,47973215,47974679,47974342,47978382,47978678,47992450,47973062,47973970,47975483,47973580,47973373,47973146],"score":105,"text":"$16","time":1777629135,"title":"Show HN: Perfect Bluetooth MIDI for Windows","type":"story"}],[47965161,{"by":"carlovalenti","descendants":6,"id":47965161,"kids":[47965791,47968399,47967932,47965994],"score":38,"time":1777567697,"title":"Show HN: TRiP – a complete transformer engine in C built from scratch just by me","type":"story","url":"https://github.com/carlovalenti/TRiP"}],[47960908,{"by":"Tdxt","descendants":6,"id":47960908,"kids":[47963312,47974504,47960923],"score":5,"text":"Two years ago I sat down to build a chess engine for my kids' Christmas present. Wanted it to play badly enough to be beatable but not randomly bad: different styles so they could choose what to face.

While building those styles I noticed the choices started looking like actual historical players' tendencies. Went looking. Turns out: given enough games from one person, you can extract enough of their decision pattern to reproduce it.

Two years later, 41 players, ~3,000 games each. Tal sacrifices unsoundly. Morphy attacks like it's 1850. Capablanca grinds endgames. Same engine underneath, visibly different choices each time.

playchessgate.com: browser, no signup. The kids never got their Christmas present.","time":1777548413,"title":"Show HN: Can a person's chess mind outlive them? An attempt with 41 masters","type":"story","url":"https://playchessgate.com/"}],[47944684,{"by":"mcohrs","descendants":0,"id":47944684,"kids":[47944745],"score":1,"text":"$17","time":1777442639,"title":"Show HN: PatchWork – A 5-pass dedup pipeline to merge N resumes into one profile","type":"story","url":"https://usepatch.work/"}],[47944584,{"by":"HatemDabet","descendants":2,"id":47944584,"kids":[47944602,47944671],"score":1,"text":"Built this because most file sharing tools either require accounts, have size limits, or store your files indefinitely. ShareNova is browser-based — live transfers are unlimited in size, no account needed for the receiver, and files aren't stored permanently on our servers. Uses WebSocket relay. Free tier available. Honest feedback welcome.","time":1777441515,"title":"Show HN: ShareNova – Browser file sharing, no account, no limits, no storage","type":"story","url":"https://sharenova.io"}],[47944106,{"by":"plonk-er","descendants":3,"id":47944106,"kids":[47946857,47944178,47944597],"score":2,"text":"Plonk 5 new places on a 3D globe every day. Test your geography, learn, and compete on the leaderboard.

I've played similar games for months and didn't like that the features I wanted were paywalled (ex: better resolution, being able to play previous days, private groups).... so I made Plonk.

Built with Globe.GL","time":1777436083,"title":"Show HN: Plonk.world – Daily Geography Game","type":"story","url":"https://plonk.world/"}],[47932531,{"by":"Naresh-cn2","descendants":1,"id":47932531,"kids":[47932673,47932536],"score":1,"time":1777371882,"title":"Show HN: Axiom v2.0 – Parallel Ingestion at 125M rows/SEC","type":"story"}],[47930665,{"by":"codesoda","descendants":5,"id":47930665,"kids":[47930983,47931487,47930728],"score":1,"text":"I'm a big user of Codex and Claude Code in the terminal. However after a big brainstorming and planning session I was finding myself with lots of comments and questions about difference places in the plan file.

So I built Discuss CLI. It turns any Markdown file (or stdin -) into a browser review surface with PR-style comments. Your existing coding agent session can gets your comments passed through to it and it reply in the margins.

This has completely transformed the way I'm interacting qith coding agents, please try it out.","time":1777353539,"title":"Show HN: Discuss CLI – No more reviewing agent plans in the terminal","type":"story","url":"https://github.com/codesoda/discuss-cli/"}],[47930504,{"by":"jkantola","descendants":1,"id":47930504,"kids":[47930635],"score":2,"text":"$18","time":1777351061,"title":"Show HN: Vaava – a baby routine tracker / logging app","type":"story","url":"https://www.vaava.app/"}],[47923015,{"by":"ioololaa","dead":true,"id":47923015,"kids":[47923104,47923159],"score":4,"time":1777304184,"type":"story"}],[47920043,{"by":"davidgomes","descendants":1,"id":47920043,"kids":[47920173],"score":4,"time":1777287849,"title":"Meta's $2B Deal to Buy AI Startup Manus Blocked by China","type":"story","url":"https://www.forbes.com/sites/siladityaray/2026/04/27/china-blocks-metas-2-billion-acquisition-of-ai-startup-manus/"}],[47920154,{"by":"comtom","descendants":0,"id":47920154,"kids":[47920155,47920159],"score":2,"time":1777288920,"title":"Electricity Provider Comparison Tool","type":"story","url":"https://elemetric.comtom.engineering/"}],[47908620,{"by":"theorchid","descendants":1,"id":47908620,"kids":[47908666],"score":2,"time":1777193456,"title":"Stripe: Analyzing first-party fraud trends","type":"story","url":"https://stripe.com/blog/analyzing-first-party-fraud-trends-account-free-trial-and-refund-abuse"}],[47908512,{"by":"lum1104","descendants":3,"id":47908512,"kids":[47908515,47910347,47908566,47923177],"score":3,"time":1777192349,"title":"Show HN: Understand Anything – a codebase knowledge graph for onboarding","type":"story","url":"https://understand-anything.com"}],[47900240,{"by":"Antibabelic","descendants":0,"id":47900240,"kids":[47900260],"score":5,"time":1777112302,"title":"Why LLMs Can't Replace Strategic Insight","type":"story","url":"https://hbr.org/2026/03/researchers-asked-llms-for-strategic-advice-they-got-trendslop-in-return"}],[47899277,{"by":"spirit23","descendants":4,"id":47899277,"kids":[47899639,47899723,47899413],"score":5,"text":"Hi everyone, I've been working on a CLI tool that can help to easily run any model in claude, Codex, Gemini, Pi, and OpenCode.

It's also an API keys manager, supports multiple providers or OpenAI/Claude/Gemini accounts. You can add openrouter, poe, Vercel AI gateways etc.

It has a built-in provider that is free to all, which is using Deepseek-V4, no login or API key required, add your own when you're ready.

After installation you can try claude instantly (No config, no login):

aivo claude

Hope it's useful to someone.","time":1777099445,"title":"Show HN: A CLI to use any model in your coding agent","type":"story","url":"https://getaivo.dev/"}],[47899375,{"by":"stealthtsdb","descendants":6,"id":47899375,"kids":[48026322,47902781,47900007,47944729,47923519,47901353,47899406],"score":11,"text":"$19","time":1777100791,"title":"Show HN: Agent MCP Studio – build multi-agent MCP systems in a browser tab","type":"story","url":"https://www.agentmcp.studio"}],[47891447,{"by":"adamewozniak","descendants":0,"id":47891447,"kids":[47891469,47899189],"score":4,"text":"$1a","time":1777043849,"title":"Show HN: TurbineFi – Build, Backtest, Deploy Prediction Market Strategies","type":"story","url":"https://www.turbinefi.com"}],[47890799,{"by":"alcazar","descendants":141,"id":47890799,"kids":[47891121,47892218,47892089,47891348,47893033,47893239,47891682,47897460,47905520,47890900,47897764,47894379,47891380,47892678,47896145,47891476,47897606,47897497,47891844,47891119,47892314,47891233,47893873,47898723,47900220,47898825,47893107,47892993,47894306,47893659,47891821,47892038,47892216,47893601,47897967,47891870,47892981,47897576,47900623,47899627,47905200,47956310,47898030,47899307,47898033,47891358,47904136,47899838,47897344,47891826],"score":532,"time":1777040932,"title":"Sabotaging projects by overthinking, scope creep, and structural diffing","type":"story","url":"https://kevinlynagh.com/newsletter/2026_04_overthinking/"}],[47891301,{"by":"stared","descendants":1,"id":47891301,"kids":[47891362],"score":5,"time":1777043260,"title":"P Is for Prefix – Vulgar Words in Polish","type":"story","url":"https://old.reddit.com/r/poland/comments/w3ip85/p_is_for_prefix/"}],[47834565,{"by":"mfiguiere","descendants":377,"id":47834565,"kids":[47836251,47835603,47839388,47835483,47834921,47837133,47834968,47836403,47842839,47835661,47848684,47838265,47840827,47835109,47837600,47845311,47837378,47835894,47843393,47836764,47844847,47840129,47838599,47835897,47839756,47844270,47844026,47835523,47838236,47845211,47837701,47838391],"score":705,"time":1776693934,"title":"Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving","type":"story","url":"https://qwen.ai/blog?id=qwen3.6-max-preview"}],[47838153,{"by":"maxfavilli","descendants":2,"id":47838153,"kids":[47838225],"score":3,"time":1776707984,"title":"Opus 4.7 – the best model nobody likes","type":"story","url":"https://maxfavilli.com/posts/opus-4-7-the-best-model-nobody-likes/"}],[47826293,{"by":"LorenDB","descendants":2,"id":47826293,"kids":[47826333,47826314],"score":5,"time":1776622167,"title":"Fixing Unix Filenames (2009)","type":"story","url":"https://dwheeler.com/essays/fixing-unix-linux-filenames.html"}],[47759035,{"by":"agentseal","descendants":27,"id":47759035,"kids":[47804379,47812549,47801125,47810979,47804428,47796590,47797785,47796907,47797603,47801532,47796769,47910159,47814544,47797590,47807216,47835564,47801485,47797936,47811536,47813575,47814769,47796297],"score":112,"text":"Built this after realizing I was spending ~$1400/week on Claude Code with almost no visibility into what was actually consuming tokens.

Tools like ccusage give a cost breakdown per model and per day, but I wanted to understand usage at the task level.

CodeBurn reads the JSONL session transcripts that Claude Code stores locally (~/.claude/projects/) and classifies each turn into 13 categories based on tool usage patterns (no LLM calls involved).

One surprising result: about 56% of my spend was on conversation turns with no tool usage. Actual coding (edits/writes) was only ~21%.

The interface is an interactive terminal UI built with Ink (React for terminals), with gradient bar charts, responsive panels, and keyboard navigation. There’s also a SwiftBar menu bar integration for macOS.

Happy to hear feedback or ideas.","time":1776121025,"title":"Show HN: CodeBurn – Analyze Claude Code token usage by task","type":"story","url":"https://github.com/AgentSeal/codeburn"}]]

user:	mockbolt
created:	April 8, 2026
karma:	4