A faster path from idea to Worker

Cloudflare Workers is designed to remove the friction between having an idea and running code at the edge. The platform handles distribution, scaling, and concurrency, so developers can focus on the application itself. Three recent updates aim to shorten that path further: a new quick-start URL, custom build support in the CLI, and a friendlier log viewer.

Start coding with a single URL

Developers can now navigate to https://workers.new to land directly in a JavaScript editor with a fresh Worker scaffold. No deployment is required to start experimenting. The editor includes a revised default template that demonstrates practical patterns such as redirecting requests, modifying headers, and parsing responses, moving beyond the usual "Hello, World!" example.

To see it in action, the short clip below shows how quickly you can jump from an idea into a runnable prototype.

Address Bar

Bring your own build for wrangler

For more involved projects—those with external libraries or written in TypeScript—the wrangler CLI remains the recommended tool. Wrangler previously supported three project types: JavaScript, Rust, and webpack. Webpack is by far the most popular due to its built-in bundling for npm packages, but developers increasingly wanted to swap in different bundlers or customise their webpack setup.

A new "custom builds" feature in wrangler (release 1.16 or later) lets you define a build script and specify the upload directory. This works across all wrangler commands, including wrangler dev and wrangler publish. With that flexibility, you can use bundlers such as Rollup, Parcel, esbuild, or your own webpack configuration.

Sample Script

Cloudflare publishes several starter templates if you want to adopt a custom build:

Full configuration details are covered in the custom build documentation.

Readable tail logs

Debugging a deployed Worker often means inspecting console.log() output or uncaught exceptions. wrangler tail streams that data live, but the output was previously raw JSON, which is useful for piping into tools like jq yet awkward to scan by eye.

The updated wrangler tail adds a "pretty" format that uses colour coding and a layout designed for terminal reading, while preserving the raw JSON mode for scripted workflows. The screenshots below show the difference between the older JSON stream and the new human-friendly output.

Sample Code

Feedback on the tail experience is welcome in the #wrangler channel on the Workers Discord server.

Where logging could go next

These updates focus on the build and deploy workflow itself, but the team hints that log visibility will keep expanding. The wrangler improvements are a step toward surfacing runtime diagnostics in other parts of the platform, so developers have more clarity when something goes wrong at the edge. Check the wrangler release notes to ensure you're on 1.16 or later to use the new features.