Last week, one of our newest additions was repo-aware settings, giving you autocomplete for your Railway config paths. This week? We're doubling down on the zero-config magic. Railway now automatically detects the environment variables your code needs and surfaces them right when you're configuring your service. 

We're also graduating two highly-requested features from Priority Boarding to general availability: HTTP metrics for better observability, and secure database access from the dashboard that routes your queries through the private network. These features have been battle-tested by our Priority Boarding users, and now everyone gets to use them.

Automatically configure services based on detected frameworks and languages

Setting up environment variables has always been one of those necessary but tedious tasks. You connect your repo, deploy your service, and then... wait, which environment variables does this thing need again? Back to the docs. Back to your code. Copy-paste. Rinse and repeat.

With Magic Config, Railway now scans your source code and automatically detects which environment variables your application is looking for. We analyze your code across 20 different programming languages, identify variable usage patterns, and surface a list of suggested variables right in your service's Variables tab — all in less than 100ms.

No more manually adding variables one by one. No more guessing which environment variables you forgot to set. Railway simply shows you what your code needs, and you can set them all at once with a few clicks.

This magic happens whenever you connect a repo to a service, and it also kicks in when you change your service's root directory. Railway stays in sync with your code, automatically updating suggestions as your project structure changes.

This is the next step in making Railway more intelligent about your code. After shipping repo-aware settings last week, Magic Config continues our mission to make Railway understand your application and automatically configure what you need.

Try it out in Priority Boarding and let us know if you run into any issues on 

What started in Priority Boarding is now available for everyone: HTTP metrics

Debugging performance issues or tracking down why your API is slow used to mean setting up external monitoring tools or instrumenting your code with logging. Not ideal when you just want a quick answer to "why is this endpoint taking so long?"

HTTP metrics solve this by giving you visibility into your service's HTTP traffic right in the Railway dashboard. 

In addition to the existing views for CPU, Memory and Public Network Traffic, you’ll be able to view the following metrics:

Response time over time, shown as latency percentiles (p50, p90, p95, p99).

Number of HTTP responses over time, broken down by status code category (2xx, 4xx, 5xx)

This way if something is going wrong, you'll be able to quickly identify the source of the issue — all without adding a single line of code to your application.

Got questions or feedback? Share them in 

Query databases securely using the private network

After launching in Priority Boarding last week, Secure Database Access is now available to all Railway users.

The Railway dashboard has always let you query databases directly, which is convenient for quick debugging and data exploration. But those queries took a detour through the public internet via your database's public endpoint — not ideal from a security or cost perspective.

With Secure Database Access, all database queries from the dashboard now flow through Railway's 

 — Queries stay inside Railway's private infrastructure without touching the public internet

 — Private network traffic means no egress fees for your database queries

 — The same dashboard experience you're used to, now with private network routing by default

Before this, if you wanted to query your database without going through the public endpoint, you'd have to either set up SSH tunneling or connect a local database client through Railway's proxy. Both options work, but they add extra steps when you just want to run a quick query.

Now, that security comes standard. Open your database in the dashboard, run your queries, and they automatically route through the private network. No setup, no configuration, just a more secure default.

If you run into any issues, let us know in 

We improved template search with weighted multi-field fuzzy search and boost scoring. Template search now prioritizes results based on relevance across title, description, category, and creator fields, with additional boosting for verified templates and popular projects. Previously, search results weren't always ordered by relevance

We fixed an issue where users could hit volume limits even when their volumes were scheduled for deletion. The volume limit check now only counts volumes with at least one active instance, preventing you from being blocked when all your volume instances are queued for deletion

We improved the visual alignment of the deploy button when viewing a service with no active deployments. Previously, the button was misaligned in the empty state — now it's properly centered

https://res.cloudinary.com/railway/image/upload/v1761340941/2025-10-24_bw4qmy.png

Magic Config, HTTP Metrics, Secure DB queries

Changelog #0261