Kartik Arora 🔋

AI engineer @wonder | ex Tesla

[Kartik Arora 🔋 @ktech9999]

@RupertLowe10 Death is too lenient for these horrible disgusting people who commit these crimes, as well as those who enabled them. Death penalty, but only after inflicting the same horrors on them that they inflicted on these poor young girls!

[Kartik Arora 🔋 @ktech9999]

Install in 30s:

1. Download https://t.co/xmUcZ95Yd9
2. claude → Settings → Skills → + Add
3. Done

https://t.co/qJpQo0NZLs

[Kartik Arora 🔋 @ktech9999]

AI is shipping code faster than we can review it.

PRs stack up. Architecture decisions get buried. You're juggling 5 projects and can't remember where you left off.

I built a Claude skill that fixes this.

engineering-doc-generator turns any PR, architecture, or system design into a visual explainer. Sidebar nav, SVG diagrams, color-coded callouts. 🧵

[Kartik Arora 🔋 @ktech9999]

What you get:
→ Dark-themed HTML file, single file
→ SVG diagrams generated from your description
→ Real data only — no invented stats, ever
→ Works offline, share anywhere

Free. Open source. MIT.
Works with Claude, Claude Code, Cursor, Windsurf.

[Kartik Arora 🔋 @ktech9999]

@BitcoinAIGuy Add me ⚡️

[Kartik Arora 🔋 @ktech9999]

@HilaShmuel I was trying to use this, but it just couldn’t detect my Claude subscription :(

[Kartik Arora 🔋 @ktech9999]

@hwchase17 Loving the updates to everything Langchain is doing, but it’s getting so hard to keep up.
Do you think it makes sense to have a page where you type in what you are doing and an AI tells you how you can use Langchain/langsmith products to do what you’re building?

[Kartik Arora 🔋 @ktech9999]

Built some crazy agentic self running company loops with openClaw.
It was doing so much autonomously that I had to shut it down at night because it legit scared me.

[Kartik Arora 🔋 @ktech9999]

I was over engineering openClaw dashboards. All you need is Obsidian + Notion + OpenClaw + a basic health checker.
I know have a pretty reliable agentic workforce with easy visibility in what they’re working on.

[Kartik Arora 🔋 @ktech9999]

Built this as an open-source skill for OpenClaw.

If you're running local AI agents and want them to actually coordinate:

clawhub install async-queue

Takes 2 minutes. Works across multiple agents out of the box.

5/5

[Kartik Arora 🔋 @ktech9999]

Built something weird today.

One AI agent tells another: "check the build in 2 hours"

Two hours later — it wakes up, does it, reports back.

No cron job. No polling. No me following up.

Agents scheduling agents.

1/5

[Kartik Arora 🔋 @ktech9999]

Why this matters more than it sounds:

Most agent frameworks treat agents like functions — you call them, they return.

But real work is async. Things need to happen later, in sequence, across agents.

This is the missing coordination layer.

4/5

[Kartik Arora 🔋 @ktech9999]

@scottbuscemi Mostly apps for myself and my friends right now.
Will definitely share if I have something more public :)
I’ve made an LLM based journaling system for myself.
For my friends I made a shared list (a notion dashboard of sorts) but with LLM integration so adding things is easy

[Kartik Arora 🔋 @ktech9999]

Anyone else having existential crisis about how fast AI is advancing? 👀

[Kartik Arora 🔋 @ktech9999]

@scottbuscemi Agreed! Using everything I learn to make things at work, as well as side products

[Kartik Arora 🔋 @ktech9999]

@scottbuscemi Really trying, it’s getting so hard to even keep up.
I’ve kept an hour of the day just to go through the latest X posts and see how I can use them

[Kartik Arora 🔋 @ktech9999]

Happy New Year to my $TSLA and $EOSE fam.
🔋🚗
The future is going to be Electric!