Vidhya Kumar

I code 🧑‍💻 and play table tennis 🏓

Building @mytoucanapp with Next.js • TanStack • Convex • Expo 🧑‍💻

[Vidhya Kumar @ByteMyTweet]

been testing OpenClaw since it dropped and yeah the security concern is real.

but the "skills fork your repo" mechanic that Karpathy mentioned? that's not a feature, that's the attack surface.

you're letting an agent rewrite your codebase to "solve config hell" but now your builds are non-deterministic and your git history is full of patches you didn't author. debugging that at 3am sounds way worse than yaml.

the tradeoff might be worth it for toys. not for anything you actually deploy.

[Andrej Karpathy @karpathy]

Bought a new Mac mini to properly tinker with claws over the weekend. The apple store person told me they are selling like hotcakes and everyone is confused :)

I'm definitely a bit sus'd to run OpenClaw specifically - giving my private data/keys to 400K lines of vibe coded monster that is being actively attacked at scale is not very appealing at all. Already seeing reports of exposed instances, RCE vulnerabilities, supply chain poisoning, malicious or compromised skills in the registry, it feels like a complete wild west and a security nightmare. But I do love the concept and I think that just like LLM agents were a new layer on top of LLMs, Claws are now a new layer on top of LLM agents, taking the orchestration, scheduling, context, tool calls and a kind of persistence to a next level.

Looking around, and given that the high level idea is clear, there are a lot of smaller Claws starting to pop out. For example, on a quick skim NanoClaw looks really interesting in that the core engine is ~4000 lines of code (fits into both my head and that of AI agents, so it feels manageable, auditable, flexible, etc.) and runs everything in containers by default. I also love their approach to configurability - it's not done via config files it's done via skills! For example, /add-telegram instructs your AI agent how to modify the actual code to integrate Telegram. I haven't come across this yet and it slightly blew my mind earlier today as a new, AI-enabled approach to preventing config mess and if-then-else monsters. Basically - the implied new meta is to write the most maximally forkable repo and then have skills that fork it into any desired more exotic configuration. Very cool.

Anyway there are many others - e.g. nanobot, zeroclaw, ironclaw, picoclaw (lol @ prefixes). There are also cloud-hosted alternatives but tbh I don't love these because it feels much harder to tinker with. In particular, local setup allows easy connection to home automation gadgets on the local network. And I don't know, there is something aesthetically pleasing about there being a physical device 'possessed' by a little ghost of a personal digital house elf.

Not 100% sure what my setup ends up looking like just yet but Claws are an awesome, exciting new layer of the AI stack.

[Vidhya Kumar @ByteMyTweet]

@claudeai but you have to keep your computer unlocked 🥲

[Vidhya Kumar @ByteMyTweet]

Chasing requests = distraction machine.
One says "dark mode." Next "export CSV." Then "AI summaries."
Week later: zero new users.
Spend that time on retention loops. Or ads. Real shit.

[Vidhya Kumar @ByteMyTweet]

Users suck at product vision.
They see your app. Think "add X." But X ain't the problem.
Your core sucks. Or pricing. Or onboarding.
Features just paint over cracks.

[Vidhya Kumar @ByteMyTweet]

Your move:

Pick the stack you know best
Ship the thing
Optimize never (until it hurts)

The market rewards speed, not cleverness

[Vidhya Kumar @ByteMyTweet]

Unpopular opinion: Your tech stack decision matters less than you think

I spent 2 weeks debating Next.js vs Remix for a side project

Meanwhile, a PHP monolith is making $105k/mo

Here is what actually moves the needle:

[Vidhya Kumar @ByteMyTweet]

The founders winning right now picked a stack and stopped thinking about it

They focused on distribution, product-market fit, and talking to users

Not whether Server Components are the future

[Vidhya Kumar @ByteMyTweet]

My niche: pediatric sleep consultants. 100 users. Growing.

Your niche is out there. It's probably "boring."

Go find it.

[Vidhya Kumar @ByteMyTweet]

The best SaaS business decision I made this year was picking a niche that makes devs yawn.

[Vidhya Kumar @ByteMyTweet]

Pick a niche where your users have money and the software sucks.

Then just... make software that doesn't suck.

Revolutionary, I know.

[Vidhya Kumar @ByteMyTweet]

@rsuyoy Lies!

[Vidhya Kumar @ByteMyTweet]

Your AI wrapper has 47 competitors. Mine has zero.

Your users churn after the free trial. Mine stick for years because their business depends on it.

Hype vs sustainability. Pick one.

[Vidhya Kumar @ByteMyTweet]

Stop building AI wrappers. Start building boring software for people who actually pay.

Here's why "unsexy" SaaS will outearn your GPT-powered side project every single time.

[Vidhya Kumar @ByteMyTweet]

Founders keep asking me "what's your moat?"

My moat is I picked a niche nobody in tech touches. My moat is I spent 6 months talking to sleep consultants instead of scrolling Twitter.

The bar is SO low in boring industries. Just show up.