Jack Herrington

. @getpostman and @tan_stack are an MCP powerhouse combo. Postman's MCP debugger and no-code MCP servers level up TanStack Start's awesome API and MCP endpoints. Check it out! youtu.be/Aaj6Ru1Lt0k

We 100% dogfood our tech on TanStack.com, but we also move *extremely* fast and are at an interesting cross section of sustainability experiments and traditional docs/open-source. We're always improving. E.g. over the last 3 days, I was able to cut total JS loaded over a session by 80%, improve all lighthouse metrics, and make the site load way snappier. You may have caught the site in the middle of some of those commits/deployments, so be sure to give it another shot!

netlify.new is live on Product Hunt. If you've tried it, or you've been watching the space, we'd love your upvote and your take! producthunt.com/products/netli…

workers-ai-provider@3.1.4 / @.cloudflare/tanstack-ai 0.1.4 - support for vision models (!) - sessionAffinity for routing multi-turn requests to the same backend - fixes for message id generation, ai gateway support

Netlify co-founder @biilmann on what's different today: Start a project from inside Netlify. Prompt, pick your agent, ship something real.

You can now start a Netlify project with a prompt. Type what you want to build, pick your agent, get a live URL on production infrastructure in minutes. This is netlify.new. Here's what that actually means: ntl.fyi/46TbFce

Really excited to join the @Netlify team and ship amazing AI and full stack features!

Community Drop [Special Announcement] x.com/i/broadcasts/1…

Welcome aboard, @jherr! If you know Jack, you know he’s been deep in @tan_stack AI, MCP, and the tooling behind AI apps that actually ship.

. @netlify's new netlify.new web app builder builds @tan_stack Start sites! Woot! Make a login and you'll get a bunch of credits to try it out for yourself. Claude Code, Codex, Gemini are all supported.

I've learnt lately that llms are definitely better at writing code than cli commands. code means type/syntax checks, can assert shapes of input/output, is super efficient (fewer tool calls etc), super portable, incredibly overrepresented in training datasets, and you can add capabilities very nicely. codemode ftw.

I recently was building with TanStack AI and just had to interview @jherr about its upcoming codemode features and how chat apps are the "hello world" of agent development! Check it out: youtube.com/watch?v=DKkE_-…

@kentcdodds @tannerlinsley @IonKorol @jamonholmgren @cursor_ai Yeah, exactly. It's really a companion. Where Claude Code always seems like... a batch job... or something. I'm super old, y'all. I did coding before the Mac and Windows were even thing. I like GUIs. I'm not nostalgic for the terminal.

@jherr @tannerlinsley @IonKorol @jamonholmgren @cursor_ai Also, any time I'm iterating on a UX thing, having browser mode and being able to just click and element and have the agent know EXACTLY what I'm talking about in the prompt is very nice.

Looking for the most reliable way to run agents in some kind of continuous/ralph loop mode, preferably overnight (@jamonholmgren). I'm use both Codex and Claude and am a sucker for OpenCode. What do I use?

@tannerlinsley @IonKorol @kentcdodds @jamonholmgren @cursor_ai system prompts and IDE integration are fantastic. It’s my #1 go to, with Claude code as a backup. I haven’t tried opencode.

@IonKorol @kentcdodds @jamonholmgren @jherr Really likes it. We should ask him.

@rauchg @cramforce @nodejs

Cloudflare’s mission is to fork the entire developer ecosystem and destroy open source. Vinext was an excuse to swindle developers into using their proprietary runtimes instead of @nodejs. It shipped with 10 vulnerabilities that they cheerfully handed to a .gov website. Now they’ve done it again. They can’t run just-bash, because it’s open source software that runs on @nodejs. It has to work “at the edge”, which is a thing no one wants (We learned the hard way… Not even the Node core developers they hired to work on Workers instead want it!)

Code mode is where its at.

@cnakazawa @wesbos @jonkkillian Got me man. Sustainable OSS funding is to multi-billion/trillion dollar question of our era. FWIW, I totally want you to win and I love Vite, Vue, etc. But I do think it's fair to look at every technology decision not just from the technology side but also sustainability.

@jherr @wesbos @jonkkillian Appreciate the feedback. What do you think we should do?

Vite just announced the Laravel/Rails for JavaScript DB, auth, KV, queues, storage, ai ++ React/Vue/Svelte/Solid and Integrates with existing metaframeworks as it's just a Vite plugin The more interesting part is they are launching their own framework with it which feels a lot like tanstack start. Has routing/loaders/actions/islands/ISR and is inspired by Inertia.js By default it's tightly bound to Cloudflare workers — which is where all the bindings come from — and will include deploying without a cloudflare account Uses better auth and drizzle under the hood

I applaud the ambition, but Void0 is another VC backed company (now in series A for $17M total). Those folks are going to be looking for a return. And that either means doing the Vercel thing and pushing services (which appears to be the play), or doing the Bun thing and getting acquired by a company. Either of which aren't that great from the consumer's standpoint.

@jonkkillian Yeah - we will see what the adapter pattern looks like, but anyone who has said next.js is lock in, will likely say the same for this

@josefbender_ @tan_stack Now and again, but so far it has always been an indication of an issue in my code.

Do you guys also get this error when hot-reoading in @tan_stack Router / Start?

@v_hadoocken @tan_stack @tannerlinsley Not planned. Just what the low hanging fruit. I would LOVE to see a PR that extended that, and in fact, you've probably just solved my e2e testing issue around this since we can just send known good/bad audio samples and get responses out of that.

@tan_stack @tannerlinsley Would you consider adding support in useRealtimeChat for audio sources other than the microphone — for example, system output / tab audio / an arbitrary MediaStream? Right now it seems the API is focused on mic capture, so I’m wondering is this planned or not.

Building realtime voice AI is usually a pain: WebRTC, audio streams, tool execution, state 😵 TanStack AI now handles that w/ useRealtimeChat()🎙️ 🔊 Voice in / Voice out 🧠 Tools/State 🖼 Multimodal ⚡ Realtime transcripts Supports OpenAI Realtime & ElevenLabs 🔗⬇🧵

@gnlzro @tannerlinsley @tan_stack When it is, we'll have some great additive additions to AI to leverage it. @tan_stack is an ecosystem!

@tannerlinsley @tan_stack RSC implementation is ready? 👀

In case you weren't looking, we just launched 4 new massive TanStack AI features today and yesterday. But our biggest one yet just got done cooking. Stay tuned. Follow @Tan_Stack to see it.