15:[[49378768,{"by":"danielvaughn","descendants":210,"id":49378768,"kids":[49380334,49380127,49379897,49379090,49379238,49380663,49380583,49384161,49384328,49380389,49386335,49384982,49380375,49382259,49383479,49381436,49381774,49379641,49379153,49379928,49380794,49382488,49384241,49385806,49380108,49387372,49386789,49379510,49394927,49380350,49380377,49380325,49387624,49385240,49380448,49380326,49380357,49380478,49381184,49383434,49380601,49379122,49380055,49380981,49381528,49387234,49379070,49379149,49380392,49383916,49382654,49380134,49386506,49382095,49380317,49380260,49380405,49380761,49380508,49380570,49387473,49379774,49385025,49380549,49379668,49380437,49381943,49385570,49382701,49380378,49381492,49380565,49386413,49382830,49381764,49380408,49386931,49379926,49380857,49380381,49379965,49380989,49398841,49380198,49380594,49379959,49380880,49380321,49380743,49380551,49380906,49382819,49380780,49380905,49380505,49418786,49393084,49408200,49381777,49384841,49385112,49385347,49385290,49382538,49379485,49391086,49392183,49385784,49380996,49385975,49383243,49384520],"score":383,"text":"$16","time":1787252736,"title":"Show HN: Huzzah – a novel approach to coding with AI","type":"story","url":"https://www.danielvaughn.dev/posts/huzzah/"}],[49373456,{"by":"simedw","descendants":118,"id":49373456,"kids":[49375955,49375352,49378199,49374380,49375748,49377213,49374339,49375833,49374197,49374623,49383322,49377995,49374300,49400045,49375371,49373700,49377302,49374316,49375900,49373738,49374420,49377970,49374395,49374249,49376572,49374793,49384033,49374355,49380651,49374837,49404181,49398867,49378917,49374384,49377781,49377036,49380875,49383703,49376725,49383191,49376750,49373678,49376365,49374571,49381911,49378775,49375713,49374340,49374948,49377832,49376258,49376158,49378080,49376221,49375187,49396690,49376286,49386112,49375272,49381061,49397428,49391111,49375239,49378721,49373681],"score":594,"text":"I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15).

The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device.

The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.","time":1787227478,"title":"Show HN: I trained a 125M model to autocomplete piano on-device","type":"story","url":"https://simedw.com/2026/08/20/midi-autocomplete/"}],[49348032,{"by":"viralpraxis","descendants":2,"id":49348032,"kids":[49350494],"score":1,"text":"Hey HN!

I'm interested in chaos engineering, automatically detecting flaky tests, and, more generally, building tools that can find/trigger invalid states in software systems.

My latest experiment is pg_disorder, a PostgreSQL extension that automatically shuffles (or reverses) the results of `SELECT` queries without an `ORDER BY` clause. While the order of rows isn't specified, applications (sometimes) implicitly assume that rows will be returned in insertion order. The project was initially aimed at finding flaky tests, but it also helped uncover bugs in projects like Gitea and Rails (see references in the repo).

Some DBMSs have this feature built-in, e.g. SQLite's `reverse_unordered_selects` pragma, but PostgreSQL doesn't.

I'd love to hear your thoughts and feedback!","time":1787069993,"title":"Show HN: Pg_disorder – shuffle or reverse unordered SELECT results","type":"story","url":"https://github.com/viralpraxis/pg_disorder"}],[49347178,{"by":"therepanic","descendants":19,"id":49347178,"kids":[49350108,49351690,49352653,49356190,49351108,49350399,49349112,49348481,49350101,49351650,49349533,49351308,49351661,49414630],"score":73,"text":"You write a standard solution, just like on LeetCode, and run it through the CLI. It identifies the problem by ID or title, executes your code against local test cases, and shows the result.

It currently supports around 1.4k problems and multiple languages, including Python, C++, Rust, Java, Go, TypeScript, Swift, and others.

The project is still an MVP. System design, SQL, and concurrency problems are not supported yet, but support for more problem types is planned.

Made in Haskell!","time":1787067040,"title":"Show HN: Openleetcode – Local LeetCode runner where tests live in the repo","type":"story","url":"https://github.com/therepanic/openleetcode"}],[49307700,{"by":"chatchan","descendants":63,"id":49307700,"kids":[49309310,49308492,49311002,49311681,49309221,49312237,49314908,49311991,49312194,49314909,49314404,49314868,49310193,49332774,49309004,49309363,49311078,49310282,49307702,49343852,49309260,49309694,49312124,49319530,49318890,49315209,49311985,49310213,49311543,49311540],"score":136,"time":1786768927,"title":"Show HN: ThoughtDAG – An editable context graph for LLM conversations","type":"story","url":"https://chenxiachan.github.io/thoughtdag/"}],[49303046,{"by":"lajosdeme","descendants":14,"id":49303046,"kids":[49309078,49303683,49303670,49310030,49306430,49316318,49304748,49379162,49347177,49310479,49306829,49310012,49312140,49315197,49313128],"score":100,"text":"Doing research with agents is fun until they blow way past budget, jumble the sources, and don't even give you the best possible answer, just sound confident.

And if you want to run some research task on local data - you have no idea where your data ends up after the prompt consumes it.

So I built this tool: a deep-research agent with an enforced budget, verified quotes, and a privacy boundary for local data.

1. Never spend more than you budgeted (measured overshoot is 0%).\n2. Every claim carries a source\n3. Data stays local (give a CSV, it'll analyze it without the data ever leaving your machine)

Works with most LLMs, including coding agents, subscriptions, local models, etc.

It's free and open source, would appreciate all feedback!","time":1786733568,"title":"Show HN: Mole – Deep research agent for your terminal","type":"story","url":"https://github.com/lajosdeme/mole"}]]

user:	gaya3bollineni
created:	August 14, 2026
karma:	2