🤨 In essence: things that worked fine in `node_modules` suddenly broke.
If you saw this warning, added an `output` path and saw errors due to that, we recommend that you simply revert to generating into `node_modules` for the time being. 🚨 But many developers hit issues after switching to a custom `output` path.
Some saw linting errors in the generated code, which broke their dev servers (because lint errors block startup).
Others ran into bundling problems, path resolution headaches, and more. 7️⃣ In the upcoming v7 major release, there'll be a breaking change: You'll need to set a custom `output` path for the generated Prisma Client instead of generating it into `node_modules`.
To help anticipate this change, we:
- added a deprecation warning
- updated our docs ⚠️ You may have seen this deprecation warning about not using a custom `output` path when using Prisma ORM lately.
We introduced this in v6.6.0 to help you prepare for a breaking change. However, the warning caused some confusion und unexpected issues.
👇 Here's what happened. 
comms from the browser company, i still like their tone and candidness. an ai-chat centric web browser to me is both horrifying and boring all at the same time, a real never-ever-do-that idea browsercompany.substack.com/p/letter-to-... 
pretty happy with this one still trying to practice daily, while walking around + living life ricoh gr3 is <3
strong cosign dansinker.com/posts/2025-0... 
What would your marketing/content team ask for if they had a magic wand?
How about the ability to build on-brand landing pages with an AI agent in minutes vs. manually reinventing the LP wheel over the course days?
We'll show you how with our new MCP this Thursday - directus.is/d7m7oIc new technology to chat with your old self: read your old journals
🎥 We're live on Youtube!
www.youtube.com/watch?v=KuU... Atuin just creeped past 24,000 GitHub stars! 🥳
ah yes, you were using gemini-4o-xx1-beta-2, of course you got bad results, you have to use openai-o3-mini-xxl and toggle on the 'do it good' button
Someone needs to interview me about this! 👀
🗳️ The results are in: you want to see Prisma Postgres in the West Coast, close to San Francisco and Los Angeles.
So `us-west-1`, here we come! 🎉
We also got a lot of love for Dallas, so we added it to our roadmap 👀
Got your next favorite region ready to go already? 
🫶
After 4 years of work, I'm incredibly excited to introduce LiveStore, the next-gen data layer I'm building for Overtone. It's based on reactive SQLite and has a built-in sync engine. Give it a try - would love your feedback!
🪐 ICYMI: Building with @astrodotbuild?
This quick guide shows how to:
✅ Add API routes to serve dynamic content
✅ Connect your Astro site to a Prisma Postgres DB
✅ Fetch and render real data on the fly
It’s fast, secure, and surprisingly:
pris.ly/guide/astro 🔔 Going live today at 5 p.m CET with VS Code!
We’ll walk through the new Prisma + GitHub Copilot Agent integration to show how you can create production ready databases & more using natural language inside VS Code.
Stay tuned!
www.youtube.com/watch?v=KuU... 
AI-generated content is creeping into bounty programs. Brace yourself for pull requests in open-source repositories.
github.com/tuist/tuist...
#oss 🤨 Prisma Client is generated into `node_modules` by default.
While this provides a familiar & unified DX, it also led to various problems.
Read our article about the original decision, what we learned since then & what changes in the future.
pris.ly/output-article 
Check out our new T-shirt design in the swag store: shop.tuist.dev/products/tu... 
Vim rebrand is crazy
"Enthusiasm" in the internet economy is a topic I've recently read about, and I wrote a small piece reflecting on it from my observations and personal experience: pepicrft.me/blog/2025/0... 🚀 The Prisma extension for GitHub Copilot is now live on the @github.com Marketplace.
You can now ask Copilot questions about your database schema, queries, or Prisma Client, right from your IDE.
🎉 No more context-switching!
Find it here → pris.ly/prisma-copilot 