Enjoying playing around with WebMCP in @OpenAI Codex to make printable guitar pick holding inserts for Altoids tins. webmcp-pick-insert.netlify.a… Deployed on @Netlify of course!
TanStack AI now has subagents 🤖
Define agents once, pass them to chat(), and each one streams its own card in your UI.
Jev routes every message to the right agents, in parallel or in order⚡
TypeSafe's Jev answers typed questions about your program state instead of writing prose. You hand it the state and declare the answers you are willing to accept. A routing question can come back as sales, support, or spam and nothing else, which means you can branch on it directly without parsing anything.
netlify.com/changelog/typesa…
I had a blast building a Fire TV Android app to run my 3D printed solar system model. Amazon has made it super easy to develop for Fire TV with React Native and Vega OS. And now they are hosting a global hackathon with great prizes. Register at fandf.co/4xqR2i5. Deadline is Oct 23, so get in soon! #sponsored @AmazonDeveloper
Gonna take some time off of X for at least the first half of September. I think it's one of several things having a negative impact on my mood so I want to see if that's the case.
Thought a few more people could use an Altoids box insert that can hold Switch games. altoids-switch-insert.netlif… Oh, and while you're there, try it out with WebMCP, it works great.
ALT Altoids box from the outside
ALT Altoids box with Nintendo Switch game holding 3D printable insert
I'm so happy people are finally realizing the beauty of WebMCP! We've been screaming "hey guys, look at this" for months lol
I guess now that more folks are using agents outside of just coding in the terminal, they see why WebMCP is sorely needed.
Here's a look into what WebMCP unlocks for agents.
aaif.io/blog/designing-websi…
Pasting every skill into the system prompt is a tax you pay on every request. 💸
withSkills() ships a one-line catalog instead. The model reads it, calls load_skill, and pulls the full instructions only when the task needs them.
Works with any provider.
Your context window is fixed. Your chat is not. 📉
TanStack AI shrinks what the model sees and leaves your transcript alone allowing for infinite compacting chats.
You pick your compaction strategy!
npm i @tanstack/ai-compaction
I'm sad to hear that Dolly Parton passed away, and so moved by how universally loved she was. she did so much good for the world, and always seemed to be a genuinely kind, happy person. RIP to a true role model
📣 WebMCP Hackathon!
If you build an innovative experience leveraging WebMCP, you can win prizes! People from across the industry are coming together for a virtual hackathon.
I’m excited to participate as one of the judges, I’ll look out for game-changing implementations.
1/2
"We blew through our annual AI budget in two months."
Chris Watts from OpenRouter talks about the next challenge for AI adoption: measuring ROI.
More code is easy to measure. Business impact is a lot harder.
After almost a year of active development, we're proud to announce that TanStack AI goes into RC! 🔥
We've done a lot of work to get it to a state we're proud of and consider stable and production ready.
Read all about it here:
tanstack.com/blog/tanstack-a…
TanStack AI now has first-party generic interrupts. 🔥
The server can stop a run and ask the user for typed data, no fake tool call, no hand-wired resume.
One definition types the payload going out and the answer coming back.
Four boundaries:
- beforeModel
- afterModel
- beforeTools
- afterTools
WebSockets just landed in TanStack AI. 🚀
One socket for the whole conversation.
Full-duplex. And if it drops mid-stream, the client reopens at the last offset and replays only the gap, no second model call.
Works on Node (`ws`) and Cloudflare (`toWebSocketResponse`)