Amin Khansari's Avatar

Amin Khansari

@akhansari.tech

Passionate about socio-technical architecture, defensive design and simple boring sustainable Ξ» code. https://akhansari.tech/ πŸ‡ͺπŸ‡ΊπŸ‡«πŸ‡·βš›οΈ πŸ—£οΈπŸ‡¬πŸ‡§

215
Followers
165
Following
267
Posts
06.08.2023
Joined
Posts Following

Latest posts by Amin Khansari @akhansari.tech

My agent skills for EffectTS, Svelte, and Tailwind
github.com/akhansari/do...

I'll add Drizzle (once V1 released) and Hono.

09.03.2026 11:06 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

My tech usage:
Before 2023: 90% Microsoft
After 2023: 10% Microsoft
Absolutely nothing is missing for me except F#.

06.03.2026 10:20 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

I did the same for Effect and Svelte/SvelteKit.
It just asked me how to split to smaller files, and done!

Amazing to see how the output quality has increased.
Much better than opinionated skills that we can found on the web.

25.02.2026 18:52 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Cool tip: You can ask LLM to build skills based on llm.txt

Like:

Build DaisyUI agent skills based on:
- daisyui.com/llms.txt
- opencode.ai/docs/skills/

25.02.2026 13:49 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

My favorite dev workflow with or without GenAI:
Function Driven Development

With GenAI:
- I model the domain with types
- I write an empty function with the desire input and output
- Either I let the LLM to complete and I review
- Or I write myself and ask GenAI to optimize and simplify.

20.02.2026 09:09 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

#opencode should be rewritten in Rust or Go.
There is too much CPU and memory usage and spikes and performance issues.

17.02.2026 09:26 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

It's super rare to see good readme files. It's an art like a good code.

16.02.2026 23:05 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Repos README:
- Before LLMs : Empty readme, 0 value
- After LLMs : 1k lines readme, 0 value

16.02.2026 12:39 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Just blogged. The slop coding era!

akhansari.tech/the-slop-cod...

12.02.2026 20:10 πŸ‘ 1 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Post image

LLM: You're absolutely correct!

07.02.2026 14:28 πŸ‘ 1 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0

Calling LLMs AI is one of the biggest marketing scams.

06.02.2026 12:38 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

people get so mad when i'm harsh on vibe coding

then they get really confused when they realize i work on opencode

which makes them even madder

01.02.2026 19:40 πŸ‘ 35 πŸ” 1 πŸ’¬ 2 πŸ“Œ 0
Preview
Agent Skills Are Spreading Hallucinated npx Commands AI agent skills are propagating hallucinated npx commands, creating real security and reliability risks for developers and supply chains.

lol www.aikido.dev/blog/agent-s...

01.02.2026 17:05 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
This is the only AI I want to use
This is the only AI I want to use YouTube video by The PrimeTime

Oh pretty cool, Primeagen has built the perfect tool for my workflow. I have to test it.

youtu.be/ws9zR-UzwTE?...

akhansari.tech/the-right-wa...

30.01.2026 20:51 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Just blogged. Mise: the underrated tool.
akhansari.tech/mise-everyth...

30.01.2026 19:06 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

So true:
"The wrong duplication is better than the wrong abstraction"

30.01.2026 12:00 πŸ‘ 3 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0

Cosmic Desktop is the future πŸ’˜

26.01.2026 07:39 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I know there are many tools around improving those. But one compiler is totally broken if it needs that to be usable.

24.01.2026 16:09 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

I think even if it's unsound there are so many opportunities for improvement.
- The speed will be ok with tsgo I guess.
- Errors are terrible. Sometimes I have 30 lines of compact text and I waste so much time trying to figure it out.
- tsconfig is hell
- multi projects repo are so complex to set up

24.01.2026 16:06 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

If you're used to using the DotNet or Rust compilers, you realize just how poor the TypeScript compiler is. It really sucks!
I hope tsgo will improve this...

23.01.2026 19:16 πŸ‘ 2 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

first we had LLMs
put it in a loop and call it an agent
put that in a loop and call it ralph

guys i think i know what's next

22.01.2026 02:40 πŸ‘ 21 πŸ” 4 πŸ’¬ 5 πŸ“Œ 0

Will current AI industries such as OpenAI, Google, Anthropic, or Nvidia become like the Big Oil (world's largest oil and gas companies) blocking the development of open-source LLM and mainstream hardwares?

20.01.2026 09:14 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Unsubscribed from Claude.

16.01.2026 23:27 πŸ‘ 3 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0
Post image

It's kind of funny that due to AI/LLMs these days another metric to rate #programming languages on has become "token efficiency".
On the bright side - some of the beloved languages like #Clojure, #Ruby, #FSharp and #OCaml fared pretty well due to their expressive and compact syntax.

16.01.2026 09:29 πŸ‘ 23 πŸ” 13 πŸ’¬ 3 πŸ“Œ 0

Oh since devstral is still in preview, it's free to use for now.
I'll give it a try.

13.01.2026 13:48 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Anthropic (Claude code) is becoming more and more evil. I'm seriously considering unsubscribing and looking for alternatives to use with opencode.
First candidate: Mistral...

13.01.2026 12:46 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

What the bullshitters on LinkedIn won't tell you, is that they vibe code a generic app without specific requirements, with premium subscriptions of 100 or 200 bucks, and they burn all the monthly available tokens in only few days...

12.01.2026 15:18 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

After a few months of using Ghostty, I go back to Wezterm.
I really like the Scrollback, Quick Select, and Copy modes.
And the fact that I can install it from Flatpak.

09.01.2026 17:31 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

In some companies there is no choice unfortunately.
Using Effet with TypeScript or using Rust are solid alternatives.

09.01.2026 16:21 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

2026 predictions
- Cursor (or another) will become the "Roblox" of software creation and distribution.
- The number and duration of big tech outages will increase significantly.
- The collapse of the AI bubble will begin with OpenAI's decline.

09.01.2026 00:05 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0