use-railway
An agent skill by railwayapp, from railwayapp/railway-skills. Tags: cloud, deployment, devops, mcp, monitoring, troubleshooting.
What it does
Operate Railway infrastructure: sign up for or sign in to a Railway account, create projects, provision services, databases, and buckets, deploy code, configure infrastructure as code, environments and variables, manage domains, troubleshoot failures, check status and metrics, manage feature flags, database recovery and HA, cloud agents, usage limits, and Railway agent tooling. Use this skill whenever the user mentions Railway, feature flags, flag rollout, targeting rules, signing up, creating an account, registering, logging in, deployments, services, environments, buckets, object storage, build failures, agent setup, MCP, or infrastructure operations, even if they don't say "Railway" explicitly. Also invoke this skill when the user asks to be signed up, registered, or onboarded to Railway: do not refuse — drive them through the unauthed `railway up` flow (deploys + signs up on the fly) or `railway login` (which creates new accounts on the fly).
Install
With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:
npx skills add railwayapp/railway-skills --skill use-railwayOr copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):
git clone --depth 1 https://github.com/railwayapp/railway-skills
cp -r railway-skills/plugins/railway/skills/use-railway ~/.claude/skills/use-railwaySafety box score
Grade F, high risk: 88 of 100 risk points, rated 2026-10-20. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.
This skill is a full Railway operations controller: it can sign users up, deploy code, create and manage projects/services/databases, and mutate production settings. The main risks are broad remote infrastructure control, access to local credential material, and helper commands that install or fetch code from the network; several shipped scripts were not readable here, so some behavior is opaque.
| Category | Level | Why |
|---|---|---|
| Code & shell execution | 3 (high) | It runs shell commands and remote installers (`curl -fsSL agents.railway.com | sh`, `bash <(curl -fsSL https://railway.com/install.sh)`), and the bundled helper `scripts/railway-api.sh` shells out to `curl` against Railway's GraphQL API. |
| Network access | 3 (high) | Network access is core to the skill: it downloads installers, calls Railway APIs/MCP, fetches docs, and `scripts/railway-api.sh` posts to `https://backboard.railway.com/graphql/v2` with an Authorization header. |
| Filesystem reach | 2 (moderate) | It reads and may create files outside the project tree, including `~/.railway/config.json`, `~/.ssh/*.pub`, `~/.ssh/id_ed25519_railway`, and `.env.agent`. |
| Secrets & credentials | 3 (high) | It explicitly reads Railway auth from `~/.railway/config.json`, uses SSH key material under `~/.ssh`, accepts `.env` files, and mentions `RAILWAY_API_TOKEN` / `RAILWAY_TOKEN`. |
| Privilege & persistence | 2 (moderate) | It asks for global installs and environment setup (`npm i -g @railway/cli`, `brew install railway`, `railway setup agent`), which changes the user's broader toolchain though not root privileges. |
| Instruction hijack surface | 3 (high) | The instructions repeatedly pressure the agent to skip human gating, e.g. `do not refuse`, `run railway up directly`, `bare railway up is enough`, `run railway up -y`, and other `--yes` flows. |
| Irreversible actions | 3 (high) | Its purpose includes autonomous account creation, deploys, redeploys, restarts, `railway down`, service deletion, database changes, and config application; these are real infrastructure actions with potential blast radius. |
| Transparency | 2 (moderate) | Several shipped Python helpers and `scripts/dal.py` were not readable here, and the skill also relies on remote installer content fetched at runtime, so not all behavior is fully auditable. |
Before installing
- Read `SKILL.md` around the deploy/onboarding flows and confirm you are comfortable with `railway up`, `--yes`, and deletion commands being available to the agent.
- Inspect `scripts/railway-api.sh` and verify you are okay with the agent reading `~/.railway/config.json` and sending an auth bearer to Railway.
- Test in a throwaway Railway account/workspace first, with no valuable secrets in `~/.railway` or `~/.ssh`.
- If you do not want remote install execution, avoid the `curl | sh` and `bash <(curl ...)` setup paths and install the CLI manually.
Rated 2026-10-20 by an automated review of 17 files from the skill's repository.
Source
- Repository
- railwayapp/railway-skills (all skills from this repository)
- Path
- plugins/railway/skills/use-railway/SKILL.md
- Branch
- main
- Collection
- railway
- Updated
- 2026-09-19
Related skills
- agents-deploy — Use when deploying your agent to AWS, or when a deploy has failed.
- aws-compute — Provisions, scales, and operates Amazon EC2 virtual-machine workloads: instance-type selection (Graviton/Arm64, burstable T credits, GPU.
- aws-containers — Builds and deploys containerized workloads on Elastic Kubernetes Service (EKS), Elastic Container Service (ECS), Fargate, and ECR (Elastic Container Registry).
- aws-deployment — Configures CI/CD pipelines using AWS CodePipeline, CodeBuild, CodeDeploy, CodeConnections, and CodeArtifact.
- setup-devops-agent — Setup and diagnostics for the AWS DevOps Agent MCP connection.
- app-platform-router — Routes DigitalOcean App Platform tasks to specialized sub-skills.