Our grade: 4 checks
Based on public data, not a full audit. See how we grade. Something changed? Request a re-grade.
Community rating
- 5 8
- 4 3
- 3 1
- 2 0
- 1 0
Trust in production
- Trust it - 83%
- With caveats - 17%
- Wouldn't - 0%
92% would recommend this server
How reviewers scored it
- Setup & docs 4.3
- Auth experience 3.9
- Reliability 4.4
- Maintainer responsiveness 4.7
Why it got an A
Railway's own MCP server now ships inside the Railway CLI (MIT, Rust, 82
contributors). A bare railway mcp is a small stdio bridge to the hosted
server at mcp.railway.com, using your railway login session, so no token
sits in your editor config. railway mcp local runs an in-process server
instead, and --oauth points your client straight at
https://mcp.railway.com, where you choose which workspaces and projects it
can reach. Railway marks destructive tools (accept-deploy,
delete-feature-flag) so clients ask before running them, and the local
server previews a change before it needs confirm: true. The hosted
server's code is not public. The old npm package @railway/mcp-server is
deprecated and its repo archived; it now only launches railway mcp.
Biggest deduction: the default connection carries your full Railway
account access and there is no read-only mode, so keep tool approval on.
How to connect
// Cursor MCP config (Railway's documented local-client config)
{ "mcpServers": { "railway": { "command": "railway",
"args": ["mcp"] } } }railway mcp install --agent claude-codeIs Railway MCP (railway mcp in the Railway CLI) safe?
Is Railway MCP (railway mcp in the Railway CLI) safe to use?
Railway MCP (railway mcp in the Railway CLI) grades A (strong public trust signals). Login is required, but access is coarse: one sign-in or key covers more than your agent needs, so grant the narrowest account you can.
Does Railway MCP (railway mcp in the Railway CLI) have known vulnerabilities?
Not in our current data: it has no open CVEs or advisories. We re-check as new ones are published.
Does Railway MCP (railway mcp in the Railway CLI) support authentication?
Yes. It requires a login, so access can be tied to a person or a scoped key.
Alternatives in Cloud & Hosting
- A AWS MCP ServerAWS's managed remote server for calling any AWS API, running sandboxed scripts, and searching AWS docs.
- A Cloudflare API MCP ServerOfficial hosted Cloudflare server that searches and calls the whole Cloudflare API (Workers, DNS, R2, D1, KV, firewall).
- A Netlify MCP ServerNetlify's official MCP server, hosted at mcp.netlify.com or run from npm, to create and deploy projects and manage env vars and secrets, forms, extensions and access controls.
- A Render MCP ServerRender's official MCP server, hosted at mcp.render.com or run locally, to create services and databases, read logs and metrics, update environment variables and trigger deploys.
- B Argo CD MCP Server (argocd-mcp)The Argo CD MCP server from argoproj-labs, run from npm or Docker, to list, create, sync and delete Argo CD applications and read resource trees, events and workload logs.
Embed this badge
[](https://trustedmcp.io/servers/railway-mcp)<a href="https://trustedmcp.io/servers/railway-mcp"><img src="https://trustedmcp.io/badge/railway-mcp.svg" alt="TrustedMCP grade A" height="80"></a>Community reviews
Solid and well behaved
Installed it on a Friday, it was part of my workflow by Monday. The OAuth flow worked first time in Claude Code. I like that access goes through a proper sign-in instead of a long-lived secret on disk. Nice that it supports both stdio and HTTP, I use stdio locally and HTTP for the shared agent. Good choice if you need Railway (railway in the Railway CLI) in your agent.
Used for Cost checks across projects
- Setup & docs 5/5
- Auth experience 4/5
- Reliability 5/5
- Maintainer responsiveness 5/5
Worth the five minutes to set up
Hooked Railway MCP (railway mcp in the Railway CLI) into Claude Code and it worked on the first try. Being official from Railway counts for a lot when I pitch it to security. Happy to run it in production.
Used for Reading logs during incidents
- Setup & docs 5/5
- Auth experience 4/5
- Reliability 4/5
- Maintainer responsiveness 5/5
Good defaults, good docs
I use this almost daily. Auth is handled properly, and revoking access later was one click. The OAuth flow worked first time in Claude Code. It is in our production agent and I sleep fine.
Used for Inventory of what is running where
- Setup & docs 5/5
- Auth experience 5/5
- Reliability 5/5
- Maintainer responsiveness 5/5
Solid core, fiddly setup
Mostly reliable, with the odd timeout. The OAuth flow worked first time in Claude Code. I like that access goes through a proper sign-in instead of a long-lived secret on disk. Fine in production if you sandbox it.
Used for Spinning up test resources
- Setup & docs 2/5
- Auth experience 2/5
- Reliability 3/5
- Maintainer responsiveness 4/5
Trustworthy enough for real work
Our agents call this dozens of times a day. Sign-in flow was painless and I never had to paste a key into a config file. Easy recommendation.
Used for Checking deploy status from chat
- Setup & docs 3/5
- Auth experience 4/5
- Reliability 4/5
- Maintainer responsiveness 5/5
Surprisingly polished
Clear tool names, sensible outputs, very few surprises. Being official from Railway counts for a lot when I pitch it to security. Updates land regularly, and nothing broke when I upgraded. Would recommend without hesitation.
Used for Reading logs during incidents
- Setup & docs 3/5
- Auth experience 4/5
- Reliability 4/5
- Maintainer responsiveness 3/5
Set it up once, forgot about it
Been running this for about a month now. Read-only mode is the first thing I enable, and it covers most of what I need. Updates land regularly, and nothing broke when I upgraded.
Used for Spinning up test resources
- Setup & docs 5/5
- Auth experience 4/5
- Reliability 4/5
- Maintainer responsiveness 4/5
Our team uses it every day
Simple to set up in Cursor, simple to reason about. Filed an issue and got a reply within two days. Scoped the token to just what the agent needs and it respected that. Nice that it supports both stdio and HTTP, I use stdio locally and HTTP for the shared agent.
Used for Cost checks across projects
- Setup & docs 4/5
- Auth experience 4/5
- Reliability 5/5
- Maintainer responsiveness 5/5
Easy win for our agent setup
Started with it as an experiment, now it is a fixture. Scoped the token to just what the agent needs and it respected that. Would recommend without hesitation.
Used for Checking deploy status from chat
- Setup & docs 5/5
- Auth experience 4/5
- Reliability 5/5
- Maintainer responsiveness 5/5
Recommended it to two teams already
Of the Railway (railway in the Railway CLI) integrations I tried, this one is the most predictable. Turned on read-only mode on day one, highly recommend doing the same. Nice that it supports both stdio and HTTP, I use stdio locally and HTTP for the shared agent. I like that access goes through a proper sign-in instead of a long-lived secret on disk.
Used for Inventory of what is running where
- Setup & docs 5/5
- Auth experience 4/5
- Reliability 5/5
- Maintainer responsiveness 5/5
Better than the DIY script it replaced
Tried it on a side project first, then rolled it out at work. Nice that it supports both stdio and HTTP, I use stdio locally and HTTP for the shared agent. The OAuth flow worked first time in Claude Code.
Used for Managing preview environments
- Setup & docs 4/5
- Auth experience 4/5
- Reliability 5/5
- Maintainer responsiveness 5/5
Saves me a context switch every hour
Set this up for our team after comparing a few options. You can tell it is actively maintained, the rough edges keep getting fixed. Nice to use something maintained by Railway rather than a random fork.
Used for Managing preview environments
- Setup & docs 5/5
- Auth experience 4/5
- Reliability 4/5
- Maintainer responsiveness 5/5
Write a review
Ran this server? Tell other teams how it went. You get one review per server. We email you a link to confirm it, then an editor checks it before it goes up. We never show your email.