Ideas Lab

Unreliable and expensive AI image generation apps

Users find Grok Imagine the worst app after a month of use, with only one successful request despite $29.95 cost and high data usage.

X · 2mo ago · Tool · AI

[FEATURE] Add the ability to manage multiple Claude accounts within the Claude Desktop app with easy switching between profiles.

### Preflight Checklist - [x] I have searched [existing requests](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20label%3Aenhancement) and this feature hasn't been requested yet - [x] This is a single feature request (not multiple features) ### Problem Statement Claude Desktop currently requires users to sign out and sign back in to switch between different Claude accounts. This creates significant friction for users who maintain separate accounts for work and personal use, or manage multiple organizational accounts. The current workflow requires me to manually sign out, re-enter credentials, and lose my current context every time I need to switch accounts. This interrupts my workflow multiple times per day and makes the desktop app less efficient than using multiple browser profiles. ### Proposed Solution Add native multi-account profile switching to Claude Desktop, similar to how Slack handles workspace switching or how Chrome manages multiple user profiles. The interface should show a profile switcher (e.g., dropdown menu in the top-right corner or keyboard shortcut like Cmd+Shift+]) that allows quick switching between accounts without signing out. There should be a command or setting that allows users to add multiple accounts, with each account maintaining: - Separate conversation history - Individual settings and preferences - Distinct MCP server configurations (if applicable) - Visual indicator showing which account is currently active ### Alternative Solutions I've tried using multiple macOS user profiles, but switching between entire user accounts is too heavyweight and interrupts my entire workflow. Currently I work around this by using Claude Desktop for my work account and claude.ai in a browser for my personal account, but this creates an inconsistent experience and defeats the purpose of having a native desktop app. Other tools like Slack, Discord, and VS Code solve this by allowing multiple workspaces/accounts within a sing

GitHub · 9mo ago · Tool · AI

A lightweight, stateless database for agent memory

I've been working on Polign and built a small prototype around something I've been thinking about with agent memory. I have built a lightweight/stateless vector db + BM25 search which works really well with typed facts and structured queries.It uses your own S3, or GCS bucket as primary storage, and restarting a node is fairly quick.Demo + writeup: https://polign.com/blog-edge-agent-memoryLive search demo: https://demo.polign.comDocs: https://polign.com

Hacker News · 2mo ago · Tool · AI

Salem Robotics (YC S26) – Software for industrial inspection robots

Hi HN, we're the founders of Salem Robotics (https://salemroboticsinc.com). We give existing mobile robots the task-specific intelligence to carry out surveys and physically interactive inspections in hazardous industrial facilities.Here's a video of it running on real robot hardware with a few words from us: https://www.youtube.com/watch?v=U_228h3NE7cWe came to Salem through robotics research at UT Austin and a combined 15 years working in nuclear, including about 10 years developing and deploying autonomous robots at Los Alamos National Laboratory. Over the last five years, we kept running into the same gap: robot hardware had become very capable, but making a robot carry out a complete industrial procedure still required a surprising amount of robotics work and manual intervention.The part that interested us most was manipulation. A nuclear contamination survey, for example, can require taking a "smear": wiping a defined area of a surface so it can be checked for removable radioactive contamination. In an oil, gas, or chemical facility, an LDAR (leak detection and repair) inspection can require moving a detector around a particular valve, flange, or connection. Other inspections require positioning an instrument at a precise location and orientation relative to a pipe or piece of equipment.These are easy tasks to compress into verbs like "wipe", "measure", or "inspect", but considerably harder to make a robot do reliably. A probe might need to remain normal to a surface throughout a path, stay within a narrow offset from a pipe, or trace a region while maintaining a particular end-effector orientation. The planner has to find a feasible motion while respecting the task geometry, manipulator kinematics, joint limits, collisions, and the environment around it.We work down to joint-level control for those interactions. One problem we've spent a lot of time on is generating constrained ma

Hacker News · 2mo ago · Tool · AI

My Claude quota ran out in 10 minutes, so I made a tool to find out why

Hacker News · 2mo ago · Tool · AI

We built open OpenRouter that turns usage into a better model

Hi HN, we built an open source model gateway. It's a single place to manage our own self hosted, frontier, and open source models in one place.It’s is rust native, built for concurrency, and implements all the config quirks across models and providers (streaming formats, tool calls, model parameters, rate limits, and different error behavior).The gateway adds under 1 ms for BYOK requests and under 2 ms when Experiential supplies the provider key. It has every major inference provider, and 1000+ models refreshed daily via a codex agent that opens a PR.Compared to other similar projects we’re open source, take no markup, allow you to mix local models with a marketplace, and use your traffic to (opt in) train you a model. Simple routing doesn’t warrant a 10% token markup.The way we do this is given standardized OTel traces, we mine representative real tasks, use text world models to simulate rollouts for various models, apply an LLM judge, and fit a nearest neighbor classifier on top of an embedding of a prompt to decide the optimal model for each request. Usually this can map out a better pareto curve on cost/quality than just calling single models but it’s not perfect.Using these simulations we can also do things like suggesting cache hit optimizations, new model suggestions, and training models.It’s open source, so you can deploy it on your own infrastructure, use our hosted version with 0 markup, or read how we design for maximum availability on our website.

Hacker News · 2mo ago · Tool · AI

I built self-hosted deployment automation tool for Windows and IIS

Seven years in the making (I admit that I did get a kid at some point which slowed down progress and brought it to a halt quite a few times), I decided to finaly release it, with a rather drastic change to the now (I believe) generous licensing model.Unlike current major players in this field, I decided to go with uncrippled, free version for commercial use for corps with up to 1M USD ARR. I poured my heart and soul into this, not because I was looking for financial success, but because I felt it just had to be done, given the price hikes wherever you look today. While my wife and kid were enjoying the beach during a vacation, I was using the time to figure out how to best tackle the problem with variable substitution and snapshots, a rather important part of the software. Pulling your hair kind of thing, many of you have likely been there, but darn does it feel good once you solve it! I know, not the best use of my time, but on the other hand, I hate the beach, so I guess this was an escape mechanism for me at that time :)I had received quite a few encouraging emails from people eager to try fDeploy back when I only had a tiny landing page out, but life has its own course and it took a while to even gather the courage to publish the first beta!As everything that sounds too good to be true in this day and age, fDeploy is no different. It does not support multi tenancy (a rather big feature which I'd love to work down the road) and a few other things may be missing which some companies may find crucial. Either way, I see this covering the needs of a large chunk of small to medium sized companies, so some "assembly" may be required. If only HN had the crowd which likes to get their hands dirty and find ways to make things work ;)I'd love to hear your thoughts and if you could take it for a spin. No signups, should be a smooth ride to get up and running and experimenting in minutes.Thanks!

Hacker News · 2mo ago · Tool · AI

Risklytics (YC S26) – Insurance brokerage for frontier tech companies

Hi HN, we're Sam and Alex, founders of Risklytics (https://risklytics.ai). We're both on leave from Harvard, and we run an insurance brokerage for companies building robots, drones, autonomous systems and satellites. You describe your machine, we find the insurance companies willing to cover it, and we read everything before you sign.Buying insurance for any frontier tech goes like this:- your first customer or pilot requires coverage before your machine can work on their site- a broker hands you an application written for ordinary businesses, and often will deny you on baseless "risks" (one of our clients got denied for using CAD software in their workflow)- you get filed under something close or outright denied, the person at the insurance company often can't tell what they're looking at, and the safe answer is no- if a quote does come back, the AI risk is often quietly removed by an add-on clause that prevents you from truly being coveredThose clauses surprised us. Insurance companies mostly don't write their own policies from scratch. They license standard forms from an industry body called ISO and attach add-on clauses to adjust them, and in January ISO published clauses that void coverage for losses arising out of AI in any part of a process. Each insurer decides whether to adopt them, account by account, and there's no public record of who has and to what capacity.There's an open question in the wording too. It defines generative AI as systems trained on data that produce text, images, audio, video or code. Whether a robot's control model, which produces motor commands, falls under that is unsettled. We've asked underwriters, the people at insurance companies who make these calls, and gotten different answers across the board.We didn't initially plan to be brokers. We came into YC to work in insurance and spent the first half of the batch on a wildfire risk model we wanted to sell to insura

Hacker News · 2mo ago · Tool · AI

TeXbrain, a LaTeX editor that runs pdfTeX in the browser via WASM

I'm a master's engineering student and a big fan of LaTeX, which I used for my thesis and research articles. I have used Overleaf and that was fine until I wanted to git sync, which unfortunately sits behind a paywall. Since I didn't want to pay subscriptions for things that should simply just work, I built the editor I wanted in my free time, where you open a tab, write LaTeX, get your PDF and the files stay all in one folder on my disk.TeXbrain is a static site with no backend. pdfTeX is compiled to WebAssembly (SwiftLaTeX) and runs in your browser. The editor can read and write your project folder through the File System Access API, so you can use git, any local TeX install, local AI Agents, or any other editor on the same files. Git is built in through isomorphic-git for anyone who would rather not touch a terminal and clone, branch, commit, push or pull via commands. No account is needed, no analytics, and everything works offline after the first load.Try it out: https://tex.swimmingbrain.dev/ (Chromium browsers get direct folder access, while Firefox or Safari fall back to a virtual filesystem)Or for self-hosting/contributing: https://github.com/swimmingbrain/texbrain (MIT, the pdfTeX engine is EPL 2.0 / GPL 2.0 and is listed in the THIRD_PARTY_LICENSES file).The part I'm most proud of getting to work is the package loading. The engine itself is only 1.8 MB. When it asks for a file it doesn't have, then a service worker intercepts the request and resolves it through Cache Storage first, then a small bundled subset, then a TeX Live tree mirrored on jsDelivr, then a SwiftLaTeX style server as a last resort. Every file is fetched once at most and after the first successful compilation, the core subset is prefetched in the background so that the offline story actually holds. Only file names go over the network, never any document content.So far, it only supports pdfTeX (no XeTeX or LuaTeX), so f

Hacker News · 2mo ago · Tool · AI

LatticeDB – Like SQLite but for graph databases

We have been using graph DBs more and more at work. I found them painful to work with locally and decided to try and build something better.

Hacker News · 2mo ago · Tool · AI

What is one simple thing LLMs are insanely bad at?

I am looking for ideas on what to train a specialized model for!What is one simple thing you repeatedly ask ChatGPT, Claude, or another model to do that it still somehow messes up?

Hacker News · 2mo ago · Tool · AI

Screen memory without screenshots, just text to Markdown

It's a macOS menu bar app that reads the text of your focused window every few seconds through the Accessibility API. No screenshots, no video, or OCR. It writes plain markdown, one file per day, into a folder you choose.Point Claude Code (or anything with file access) at the folder and ask "what did I work on Tuesday?" or have it build memory about your projects. The folder contains an AGENTS.md explaining the format to the model.Thought it might be useful for someone so chucked it up on GitHub

Hacker News · 2mo ago · Tool · AI

OzBrain, a shared brain for knowledge between agents and your team

I think agent-first chat interfaces will be a primary software modality and busy dashboard/UI will go away. I’m not sure who exactly wins it, but I want my knowledge to grow/go with me.A lot of the “knowledge” ie research, analysis, reasoning will be done by agents as the primary user. Our current notes tools & tasks management systems were built for humans… I don’t care what the 17th thing on my bug backlog is. I want to conduct agents that can execute for me and do great work.What I built OzBrain to do: + Create a central place for agent reasoned knowledge to live + Be agnostic about what apps/agents connect to it + Capture everything and track it so I can audit it + Enable teams, collaborators or partners to share brains + Handle conflicts so many agents in the same article doesn’t blow up + Refactor knowledge into more token friendly chunks and map the index well + Close the knowledge loop so new thinking supersedes old thinking across the corpus. Don’t erase, depreciate and link + Keep user data safe and secure ++ Be easy enough to use that you don’t have to have any technical knowledgeSome among us will always build their own custom solutions, but there are millions of tech professionals and small business owners that will use agents heavily and need a solution. So I’m trying to build that.Isn’t this like gBrain? Yes, similar. I think it’s like AWS vs Vercel. AWS is very powerful, configurable, and useful if you’re technical and want to invest the time into really fine tuning your system… but if you just want your web deploy/hosting to just work and be easy to deal with you use Vercel.// WHY I MADE ITI’ve been enjoying getting back to my technical roots, as I lost my coding skills more than a decade ago, but with AI I can focus on the system and the product in partnership with agent coding workflows.I recently built a Voice AI for older people. To build it I created an agentic engineering workflow (feel free to rip that up as

Hacker News · 2mo ago · Tool · AI

Lack of curated, high-quality Apple Shortcuts for AI automation and workflows

The Shortcuts Gallery is filled with outdated or low-quality user-created automations, making it hard for users to discover effective AI and app integration workflows.

X · 2mo ago · Tool · AI

Huzzah – a novel approach to coding with AI

Hello everyone. I've been working on this experimental editor called Huzzah.I've been working almost exclusively with coding agents since January of this year, and over the past few months I began to feel utterly exhausted by them. They're great, but I'm finding it more and more tedious to write full sentences for every change I want. Not only that, but it seems there's a complexity limit for codebases - beyond a certain point the agent begins confusing itself.I'd like to go back to writing code, but I don't want to go all the way back to fully manual coding. So I've come up with this interaction paradigm where you: 1. write pseudocode in whatever way makes the most sense to you 2. on save, the editor synchronizes your work to real source code 3. the pseudocode is persisted alongside the generated code, making your prompt effectively a stored record of intent. It may not work for every use case, but in my initial playthroughs I've found it very enjoyable.Right now it's just a proof of concept - installation instructions are here in the readme: https://github.com/danielvaughn/hzYou can also watch a video of it in action here: https://x.com/danielvaughn/status/2090456808431165715Cheers!

Hacker News · 2mo ago · Tool · AI

Make your logo extra bright on HDR screens

Certain logos started standing out to me on LinkedIn as brighter/whiter than everything else around them.I dug in and found out this is accomplished by adding a gain-map to an existing JPEG, visible only on HDR screens like a newer MacBook Pro. LinkedIn is the only social network I've found that isn't stripping them out, but of course you serve them up on your own site.I worked with Claude Code to turn it into a little browser-based utility (no registration) and hope you find it useful!

Hacker News · 2mo ago · Tool · AI

Proliferate- open-source, self-hostable Codex for any coding agent

Hi HN- I'm Pablo, the founder of Proliferate (YC S25)!Proliferate (https://github.com/proliferate-ai/proliferate) is an open-source, self-hostable AI IDE that lets you work and automate tasks with Claude Code, Codex, OpenCode, Cursor, and Grok in one place.Here's a quick 2m demo of how we use Proliferate to build Proliferate: https://www.youtube.com/watch?v=tGNX0oaWmBYI started building Proliferate after my team onboarded to OpenAI Codex. Within days, we were using it for everything: using computer use instead of navigating websites ourselves, having Codex coordinate other agents, and setting up automations for recurring work. We really never needed to leave the desktop app to get work done.If my team’s experience is anything close to representative, a Codex-like app (a horizontal agent with a beautiful UI) is the main interface every company is going to use to get work done. That is perfectly in line with OpenAI’s mission to make Codex the everything app (see: https://news.ycombinator.com/item?id=47796469).But as we started automating work closer to the core of the business, I wanted to work with agents from all the labs, including open-weight models, without becoming increasingly dependent on OpenAI.And that’s what Proliferate is for! It's the open-source, self-hostable Codex that preserves your optionality across agents and model providers while building toward Codex’s breadth.Today Proliferate supports:* Working with Claude Code, Codex, OpenCode, Cursor, and Grok with their native inference and configuration options, including Bedrock, Azure, and self hosted inference.* Inter-agent communication and management: a parent agent can spawn and communicate with another supported agent as a subagent (I personally like to have Fable delegate to Codex, with OpenCode models reviewing PRs).* Building workflows- one of the features I'm most excited about. These are re-usable chains of agent sessions and

Hacker News · 2mo ago · Tool · AI

Omacosy – Omarchy-style tiling desktop for macOS, no SIP

I have been using omarchy on my tower since nearly a year now, shortly after it was released first. I really love the experience I am having with it but I still use my macbook for daly work, so I wanted to recreate a similar experience on it. Thats why I created omacosy, a setup for tiling windows, custom menu bar, some themes from omarchy, focus follows mouse, focus rings around windwos, some mac flavors with trackpad events and a custom mission control overview for your workspaces.I used AeroSpace over yabai for the tiling window manager because I didnt wanted to compromise on SIP which is a mac security feature. It is supposed to be keyboard first like omarchy to move windows organize workspaces etc The setup runs around 157mb of ram and consists of AeroSpace, Karabiner (for the super key), and five small self build swift binaries.I am running it daily on my M1 max macbook, currently on macOS26. I havent tested it much on other macbooks or macOS versions. The install script creates a manifest file to backup what was installed before and what it installed itself, the uninstall script takes that into account to clean up the macbook to exactly the state it was in before. It needs quite some permissions for it sfunctionality which I layed our in the project readme. I wanted to be really transparent about which permissions it uses and for what reason.I would love to get some feedback or see people trying it out and hearing your opinion. Mostly about what still doesnt feel smooth in the experience or if you find any performance issues.

Hacker News · 2mo ago · Tool · AI

Anonymous age verification with passkey-powered encryption

This project has been kicking around in my head since I first heard about the webauthn PRF extension in early 2024. I've slowly chipped away at it since, and finally got things to a shareable state over the summer thanks to a very fun parental leave. Headed back to work tomorrow, so I figure there's no time like the present.tldr: A client-held encryption key derived during passkey authentication encrypts all sensitive user data prior to persistence so that only the user is able to decrypt and reuse that data on their device. This allows short-lived, privacy-preserving age proofs to be issued to requesting applications (18+, no PII shared) without requiring users to re-upload their documents. The SSO user experience is built on top of the OAuth 2.0 Authorization Code Flow.https://loginwithone.com - demo video + high-level architecture + FAQI also made the parody demo apps from the video public if anyone wants to play around with the user experience:https://demo.brainrot.loginwithone.com https://demo.dgnrt.loginwithone.com https://demo.kirby.loginwithone.comI suspect most will choose to pass on the ID stage for now (no offense taken, doing so is low reward in this context) but if you navigate to https://app.loginwithone.com after onboarding you can demonstrate the passkey-powered encryption on your email via the lock/unlock button.Very open to feedback and happy to answer any questions! I plan to pull the client-side encryption functionality into an open-source typescript library for general use, so any thoughts or suggestions on what you’d like to see out of that interface would be supremely useful. Thanks all,Michael

Hacker News · 2mo ago · Tool · AI

I trained a 125M model to autocomplete piano on-device

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.

Hacker News · 2mo ago · Tool · AI