#builds

Ryu Hi Shirazatz, Your repro is already pretty strong, especially because an old green commit fails only after the CLI version changed. One workaround I’d try is making the function pattern less dependent on the exact file path: or, if Vercel accepts the JS glob in this setup: The error is coming from the functions pattern check, not from your function code, so this test should tell you whether the regression is specifically around matching api/index.js under a custom Root Directory. I’d also include a minimal public repro if possible: with Root Directory set to app and only the functions config n

Ryu Hi Brian, Since manual redeploys work but Git pushes and Deploy Hooks create no visible deployments, I’d check the settings that can stop deployment creation before digging into build logs. A few quick checks: I’d also search the repo for Git deployment-disabling config: If you have either of these, they can prevent Git-triggered deployments: or the older form: One more useful separation test: create a brand new Deploy Hook for the exact branch you expect, then trigger that new hook once with curl. If the new hook also returns PENDING but no deployment appears, and Git pushes also create nothi

Ryu Hi Adeline, Your suspicion sounds pretty reasonable. When an AI app grows from "one prompt generates everything" into multiple muses, menus, timelines, variations, etc., the fragile part is usually that too much decision-making is left inside the prompt. I'd separate this into a more deterministic pipeline: The biggest fix I'd make is: don't ask the model to both choose the correct muse and generate the entire party plan in one large prompt. Pick the muse in code first, then pass only that muse's allowed values into the generation step. For example: For the repeated / drifting output, I'd also

Vinaydabas611 40 viewers/day is low for 3h50 of CPU, so before pinning it on crawlers or compute, confirm your actual request count in the logs. If it's far above your real traffic, the usual hidden cause is your own app re-firing server renders --- a router.refresh() / auto-refresh on tab focus or navigation, aggressive client-side polling, or default prefetch firing a burst of RSC requests on every page. (A third-party uptime monitor on a tight interval does the same.) The tell is a repeating, multi-route request pattern in the logs that no human would produce. If real crawlers are a big share: a robots.t

Don Lao problem is resolved by remote supabase integration, use API or token instead.

system The domain troubleshooting guide can help with most custom domain configuration issues. You might be able to use that guide to solve it before a human is available to help you. Then you can come back here and share the answer for bonus points. vercel.com/docs/projects/domains/troubleshooting You can also use v0 to narrow down the possibilities.

js I'm seeing the same problem on one of my three projects. I've tried both Chrome, incognito Chrome, and Firefox, all with the same issue. I'm on a mac. Nextjs as well.

js And now it's fixed.

Muhehee Bumping this for the dev team's visibility. Since support confirmed this is an infrastructure bug but Hobby users can't open tickets, we really need an engineer to check this out. Our staging environment is still blocked.

Pauline P. Narvas It sounds like there is no Next.js version in your package.json file. Can you check for this? How did you do this? 🙂

Ryu Hi, That error usually means Vercel cannot find the package.json that contains next, or the project root is set incorrectly. A few things to check: In your repo, confirm next is listed in dependencies or devDependencies in the same package.json Vercel is building from. In Vercel, open Project Settings -> General -> Root Directory and make sure it points to the folder that contains that package.json. If this is a monorepo, set the root directory to the actual app folder, for example apps/web. Make sure the repo includes the lockfile from the same package manager you are using. Redeploy after ch

Pauline P. Narvas Hey, @suheb! Feel free to share more details about your project, it'd help us tailor our suggestions, but some ideas to get you started: Bundle Size Optimization Use <@next> /bundle-analyzer to identify large dependencies: ANALYZE=true npm run build Enable optimizePackageImports in next.config.js for icon/utility libraries Implement code splitting and lazy loading for non-critical components Tree-shake unused code and remove unnecessary dependencies Image Optimization Use Next.js <Image> component for automatic optimization Images are served in modern formats (WebP/AVIF) reducing size by 25-35

Ryu A good way to approach this is to separate build/deployment size from runtime loading performance. For deployment/build size: Remove unused dependencies. Check for large assets accidentally included in the repo. Avoid bundling server-only packages into client components. Use dynamic imports for heavy UI/client-only code. Keep generated files, uploads, and local cache folders out of the deployment. For page performance: Use next/image for images. Compress large images before uploading them. Lazy-load below-the-fold components where it makes sense. Check bundle size with a bundle analyzer. Use V

Pauline P. Narvas Hey, Brightpixels! Welcome to the Vercel Community 👋 For anyone else reading this, this is about the live property from the Vercel REST API, docs here: vercel.com/docs/rest-api/projects/retrieve-a-list-of-projects From my understanding, the live property appears to be an internal state that tracks whether your production deployment is actively serving traffic, which is separate from the paused state. When a project is paused, both paused: true and live: false are set. It looks like the unpause endpoint might not be properly restoring the live state in your case. To resolve this: Try triggering

sprucepad The title seems to have been changed, it shows “25m / 0s” on the dashboard. It also seems to be completely ignored, as I just deployed something just fine.

sprucepad It seems to have disappeared, so maybe Vercel’s adding build time limits and it was glitched out.

Aconfusedprogrammer I had this issue too. I still have it, but I can still deploy just fine.