raunaqvaisoha
raunaqvaisoha on Hacker News
Posted by
raunaqvaisoha on Tuesday April 14, 2026 @04:10PM
from the show-hn dept.
https://github.com/crocsarecool/nutrition-app
I was tempted by Cal AI, but noticed the camera action was a bit awkward and high effort for me, as someone who doesn't take a lot of photos and also remembers to log stuff after I've already eaten it.So I built a simple app to talk to your calorie note taker, who puts a simple MD diary and also consults with you on your diet strategy.I simply eat and then tell this app what I ate, it uses...
Posted by
raunaqvaisoha on Friday March 13, 2026 @08:51AM
from the show-hn dept.
https://github.com/vibetrade-ai/vibe-trade
Post-Opus 4.6, LLMs feel much better at using bash, code, local files, and tools.So I kept coming back to a simple question: if a model can use a computer reasonably well, why can’t I just give it my broker account, a strategy, and let it trade?My conclusion is that the blocker is not model capability in the abstract. It is the system around the model.A raw LLM breaks on a few practical things...
I feel like a lot of folks are saying this kills the Code on your Phone opportunity some start-ups are building for. I don't agree. I feel like coding agents are like streaming services, we will subscribe to multiple and switch between them. So for one there's value in a universal control plane. The other is that mobile as a coding interface should offer more than a remote control to the desktop. I think there's still some space to cook, especially if people are investing 8 hours a day talking to agents, the interface surely matters.
Posted by
raunaqvaisoha on Monday February 23, 2026 @05:47AM
from the self-post dept.
I want Claude to help me manage my money. I'm thinking of forking OpenClaw and giving it data and broker access, has anyone tried doing this, what do you think?
I am trying to document learnings from debug sessions in an MD file as a sort of runbook. Looking for skills that could make that simpler.
Posted by
raunaqvaisoha on Wednesday February 04, 2026 @01:26PM
from the nathanonn.com dept.
Focus is a currency and you have a limited amount of it, if all SaaS is built internally, teams would go bankrupt. There's likely always going to be a band of experts focused on solving a problem and everyone pays them to solve it for them, because they do it better and can handle the hassle of maintaining it.
I like the idea of saving recipes online. What I wish existed was something which would let me take a photo of a dish I’m eating and tell me how to replicate it at home. There’s been so many times I’ve eaten something at a restaurant or someone’s house and want to make it at home, but it’s inconvenient to ask for the recipe.
That's what I meant to say, that task assignment to agents could shift to mobile. What experience would you personally like, from a mobile continuity perspective?
Interesting, what got you interested in having a mobile setup btw?
> Did you have something specific in mind?
I'm still brainstorming on this topic, would love any ideas. But one habit I am noticing amongst my friends and I is assigning a task to an agent, then moving onto the next task while the agent works on the previous one (this can be in codex, claude etc). All of this doesn't need to be on the laptop, infact it could be done through whatsapp or telegram. Let's say I have an ongoing chat with the agent, me giving instructions, getting a diff, reviewing and then hit to commit. Would be useful when I am away from my desk.
Another thing could be an information assistant, to keep me in the loop with what my agents are upto by summarising their performance and current active tasks and letting me butt in when they're stuck somewhere.
As you can tell I am still thinking it through :)
Me too, I find myself gaining or loosing motivation for projects based on social sentiment online. When you're in the real world, it's so different almost as if it's completely disconnected from what we see online. Especially in tech waves, we tend to focus on the fringe instead of making basic progress because we're afraid we're already too late. But when you step outside and talk to real people, you realise it's really not too late. So I agree, ideas get better the further away you are from a computer :)
Interesting! I've been wanting to build something in this space. Any learnings you can share about your experience so far?
I've been thinking with the advancements of agents being more capable of being left alone and just treated like a co-worker, is anyone actually finding themselves doing the assingment tasks from their phone, away from their laptop.
Posted by
raunaqvaisoha on Tuesday February 03, 2026 @06:50AM
from the self-post dept.
Posted by
raunaqvaisoha on Friday July 18, 2025 @10:28AM
from the show-hn dept.
https://www.npmjs.com/package/convo-sdk
Hey HN!I built Convo SDK after spending way too many hours wrestling with LangGraph checkpointer implementations. Every time I wanted to add memory to a LangGraph agent, I'd end up in database setup hell.The Problem:
- Setting up PostgresSaver means managing connection pools, schemas, migrations
- MongoSaver requires MongoDB hosting, backup strategies, monitoring
- Custom BaseCheckpointer...
Posted by
raunaqvaisoha on Tuesday July 15, 2025 @05:53AM
from the chatgpt.com dept.
Posted by
raunaqvaisoha on Wednesday July 09, 2025 @07:33AM
from the show-hn dept.
https://www.bookshelf.diy/
I’ve been hacking on a project called Bookshelf (https://www.bookshelf.diy/). It lets you take an archive — say, your Substack export, a bunch of PDFs, or even saved HTML files — and turn that into a retrieval-backed GPT that your readers can query.The idea is: instead of scrolling archives, they just ask questions. Answers are pulled only from your original content, with citations.It’s aimed...