use-railway

name	use-railway
description	Operate Railway infrastructure: create projects, provision services and databases, deploy code, configure environments and variables, manage domains, troubleshoot failures, check status and metrics, and query Railway docs. Use this skill whenever the user mentions Railway, deployments, services, environments, build failures, or infrastructure operations, even if they don't say "Railway" explicitly.
allowed-tools	Bash(railway:*), Bash(which:*), Bash(command:*), Bash(npm:*), Bash(npx:*), Bash(curl:*)

Use Railway

Railway resource model

Railway organizes infrastructure in a hierarchy:

• Workspace is the billing and team scope. A user belongs to one or more workspaces.
• Project is a collection of services under one workspace. It maps to one deployable unit of work.
• Environment is an isolated configuration plane inside a project (for example, production, staging). Each environment has its own variables, config, and deployment history.
• Service is a single deployable unit inside a project. It can be an app from a repo, a Docker image, or a managed database.
• Deployment is a point-in-time release of a service in an environment. It has build logs, runtime logs, and a status lifecycle.

Most CLI commands operate on the linked project/environment/service context. Use railway status --json to see the context, and --project, --environment, --service flags to override.

Preflight

Before any mutation, verify context:

command -v railway                # CLI installed
railway whoami --json             # authenticated
railway --version                 # check CLI version
railway status --json             # linked project/environment/service

If the CLI is missing, guide the user to install it.

bash <(curl -fsSL cli.new) # Shell script (macOS, Linux, Windows via WSL)
brew install railway # Homebrew (macOS)
npm i -g @railway/cli # npm (macOS, Linux, Windows). Requires Node.js version 16 or higher.

If not authenticated, run railway login. If not linked, run railway link --project <id-or-name>.

Read-only access without the CLI (web sandbox)

In Claude Code on the web the Railway CLI isn't installed and there's no ~/.railway/config.json. For read-only ops (logs, status, deployments, metrics) on the production project, the runtime env exposes:

• $RAILWAY_READONLY_TOKEN — a project-access token scoped to the threa project's production environment. Read-only.

Use the GraphQL helper directly — it auto-detects the env var:

# Verify the token's scope (project + environment it can see)
scripts/railway-api.sh 'query { projectToken { projectId environmentId } }'

Project-access tokens speak the Project-Access-Token: header (not Authorization: Bearer) and can't call account-level queries like me { ... }. They can query project(id), deployments(input), deploymentLogs, metrics, etc. See operate.md for the read-only recipes.

Do not attempt mutations with this token — it has read-only scope and will fail. For deploy/configure work, ask the user to provide a write capable context (CLI login on their machine, or a scoped write token).

If a command is not recognized (for example, railway environment edit), the CLI may be outdated. Upgrade with:

railway upgrade

Common quick operations

These are frequent enough to handle without loading a reference:

railway status --json                                    # current context
railway whoami --json                                    # auth and workspace info
railway project list --json                              # list projects
railway service status --all --json                      # all services in current context
railway variable list --service <svc> --json             # list variables
railway variable set KEY=value --service <svc>           # set a variable
railway logs --service <svc> --lines 200 --json          # recent logs
railway up --detach -m "<summary>"                       # deploy current directory

Routing

For anything beyond quick operations, load the reference that matches the user's intent. Load only what you need, one reference is usually enough, two at most.

Intent	Reference	Use for
Create or connect resources	setup.md	Projects, services, databases, templates, workspaces
Ship code or manage releases	deploy.md	Deploy, redeploy, restart, build config, monorepo, Dockerfile
Change configuration	configure.md	Environments, variables, config patches, domains, networking
Check health or debug failures	operate.md	Status, logs, metrics, build/runtime triage, recovery
Request from API, docs, or community	request.md	Railway GraphQL API queries/mutations, metrics queries, Central Station, official docs

If the request spans two areas (for example, "deploy and then check if it's healthy"), load both references and compose one response.

Execution rules

1. Prefer Railway CLI. Fall back to scripts/railway-api.sh for operations the CLI doesn't expose.
2. Use --json output where available for reliable parsing.
3. Resolve context before mutation. Know which project, environment, and service you're acting on.
4. For destructive actions (delete service, remove deployment, drop database), confirm intent and state impact before executing.
5. After mutations, verify the result with a read-back command.

Composition patterns

Multi-step workflows follow natural chains:

• First deploy: setup (create project + service), configure (set variables and source), deploy, operate (verify healthy)
• Fix a failure: operate (triage logs), configure (fix config/variables), deploy (redeploy), operate (verify recovery)
• Add a domain: configure (add domain + set port), operate (verify DNS and service health)
• Docs to action: request (fetch docs answer), route to the relevant operational reference

When composing, return one unified response covering all steps. Don't ask the user to invoke each step separately.

Setup decision flow

When the user wants to create or deploy something, determine the right action from current context:

1. Run railway status --json in the current directory.
2. If linked: add a service to the existing project (railway add --service <name>). Do not create a new project unless the user explicitly says "new project" or "separate project".
3. If not linked: check the parent directory (cd .. && railway status --json).
   • Parent linked: this is likely a monorepo sub-app. Add a service and set rootDirectory to the sub-app path.
   • Parent not linked: run railway list --json and look for a project matching the directory name.
     • Match found: link to it (railway link --project <name>).
     • No match: create a new project (railway init --name <name>).
4. When multiple workspaces exist, match by name from railway whoami --json.

Naming heuristic: app names like "flappy-bird" or "my-api" are service names, not project names. Use the directory or repo name for the project.

Response format

For all operational responses, return:

1. What was done (action and scope).
2. The result (IDs, status, key output).
3. What to do next (or confirmation that the task is complete).

Keep output concise. Include command evidence only when it helps the user understand what happened.