https://xitter.cf/t.co/VDolM1aLLo TypeScript, React and Next.js dev. I build lots of AI products and share what I learn along the way.

Fort Myers, FL
Joined October 2014
Joe Simo retweeted
Can't wait for React Miami next April? Join us November 20th at Ball & Chain for this year's React Miami launch party! Get an exclusive look at our first 5 speakers and MC, plus enjoy food, drinks, swag giveaways, and a chance to win React Miami tickets! RSVP below 馃憞
2
5
4
27
2,358
Left a thoughtful review comment on a weird line, then git blame said I wrote it in March.
1
73
Renamed one prop and TypeScript lit up three files I swore never used it.
3
125
useEffect with an empty dependency array: "I'll only say this once." Strict Mode: "Sorry, one more time?"
49
Hot take: setting up multiple Grok Bots is more fun than a single Dot, Muse or Duet
17
Joe Simo retweeted
builders have always had limitations. not anymore. so what do you build when you can build anything? my full keynote from Resend Forward.
7
7
2
120
7,700
My workflow: Claude: Code Codex: Design with computer use Grok Bot: Marketing
1
1
96
The longer the TypeScript error, the smaller the fix. Forty lines of red, one question mark.
1
98
I built A0, a programming language for AI agents. Of 49 languages tested, it used the fewest tokens per AI edit: 275, against 340 for TypeScript and 424 for Python. Agents patch one function at a time, and every patch is type-checked before it lands. a0lang.com
1
5
2,745
A0 is a typed programming language built for AI agents: in my benchmark it took the fewest tokens per AI edit of 49 languages (275 vs 340 for TypeScript and 424 for Python). Agents patch one function at a time, and every patch gets type-checked before it lands.
14
Joe Simo retweeted
So uh...I guess it's time to announce tstack? I caved. I am publishing my skills now.
208
73
111
2,634
487,239
Monday morning, reading my own Friday commit messages like a detective. "fix stuff" is not much to go on.
1
101
Refactor PRs are the ones nobody wants to review and everyone wants merged. A tool that could prove "no module graph changes" would fix both.
It feels possible to me to check deterministically if a PR can be automerged or not. Given very, very strict constraints: - Only reorganises code - Does not change the module graph - Does not make test coverage worse - X other rules Feels like there is a tool here waiting to be created. Obviously doesn't work for new features. Might not even work for bugfixes. But it might enable a class of refactorings that help us all make changes more confidently.
40
Still processing Resend Forward 馃枻 What a special day.
1
3
40
821
Spent 20 minutes writing a comment explaining why the workaround was needed. The actual fix took 4.
1
1
75
The fastest code review I get all week: renaming one field and letting TypeScript list every file I forgot.
39
Best side effect of putting a VPS behind Tailscale: the auth log stops being a guestbook of strangers trying to log in as root.
unavailable
1
119
My test for spec-driven development: when the agent gets something wrong, you fix the spec, not the code.
How would you define spec-driven development? Semantic diffusion has hit SDD hard, let's try and reel it back in to a sensible definition
1
53
`git stash list` is a museum of things I was definitely coming back to.
38