use-railway_skill

Overview

This skill operates Railway infrastructure from the CLI and API: create projects, provision services and databases, manage object storage buckets, deploy code, configure environments and variables, manage domains, and troubleshoot builds and runtime issues. It guides safe context checks, uses JSON output for reliable parsing, and follows a clear preflight and verification flow. Use it whenever you mention Railway, deployments, environments, buckets, or infra operations.

How this skill works

The skill inspects the current Railway context (workspace, project, environment, service) using railway status --json and other CLI commands, then chooses the appropriate action (create, configure, deploy, or debug). It prefers the Railway CLI, falling back to an internal API script for unsupported actions, and always confirms context before mutating resources. After changes it performs read-back verification and returns concise actionable results including commands used and next steps.

When to use it

• Create or link a project and add services from your repo or monorepo.
• Deploy code, redeploy, or restart services and check deployment status and logs.
• Provision or inspect managed databases and S3-compatible buckets and credentials.
• Set, list, or update environment variables and environment-specific config.
• Triage build or runtime failures, check metrics, and verify recovery.

Best practices

• Always run railway status --json and railway whoami --json before mutations to resolve context.
• Use --json flags for deterministic parsing and scripts; prefer CLI over ad-hoc API calls.
• Confirm destructive actions explicitly and show expected impact before proceeding.
• After mutations, run a read-back command (e.g., railway service status --json) to verify success.
• When adding services in a monorepo, set rootDirectory and avoid creating a new project unless requested.

Example use cases

• Link the current repo to an existing Railway project and add a new service for a sub-app.
• Deploy a new release with railway up --detach -m "message" and then fetch recent logs to confirm health.
• Create an S3-compatible bucket, retrieve credentials, and configure them as service variables.
• Investigate a failed build by collecting build logs, identifying the error, applying a config fix, and redeploying.
• Add a domain, map it to a service port, and validate DNS propagation and service health.