MCP Server

railguey

Railway deployment management from Claude Code. Deploy, rollback, logs, variables.

pip install railguey

View on GitHub 1 stars 954 downloads/mo v0.3.1 May 2, 2026

What it does

railguey is an MCP server that gives AI agents full control over Railway deployments. Create projects, deploy services, check logs, set variables, roll back — all through tool calls without leaving your terminal.

What the flow looks like

An agent shouldn't need a dashboard to operate production. Railguey exposes Railway's lifecycle — status, logs, deploys, rollbacks, variables — as MCP tools. When a deploy misbehaves, the agent sees it, rolls it back, and confirms health without leaving the conversation.

1. Last deploy is leaking 500s — agent pulls HTTP logs

agent · production

  › railguey status ~/apps/web service web · deployment d2f9 (5m ago) · DEGRADED http p50 120ms · p95 4.2s · 500s: 8.1%  › railguey http_logs web --status 500 --lines 5 20:14:02 POST /api/checkout 500 "TypeError: Cannot read..." 20:14:07 POST /api/checkout 500 "TypeError: Cannot read..." 20:14:15 POST /api/checkout 500 "TypeError: Cannot read..."  agent same stack every time → rollback.   

2. Rollback + confirm — no dashboard needed

agent · production (rollback)

  › railguey rollback web --to d8a1 railguey d2f9 → d8a1 · deploying... railguey d8a1 LIVE (22s)  › railguey status ~/apps/web service web · deployment d8a1 (30s ago) · HEALTHY http p50 110ms · p95 380ms · 500s: 0.0%  agent filing issue with the stack trace; d2f9 stays parked.   

Same operator flow a human would run in the Railway dashboard, expressed as tool calls the agent can chain. 17 tools across status, logs, deploys, variables, domains, volumes, and doctor audits.

Key features

Multi-account — manage multiple Railway accounts and workspaces
Full lifecycle — create projects, services, environments, domains
Deploy + rollback — push code and roll back if it breaks
Logs — build logs, deploy logs, HTTP logs with server-side filtering
Doctor — workspace, service, and project-level health audits
Variables — set env vars without the dashboard
TOTP — built-in 2FA for Railway operations that need it

Install

pip install railguey

Usage

# CLI
railguey status ~/repos/my-app
railguey logs ~/repos/my-app web --lines 50
railguey deploy ~/repos/my-app web
railguey doctor ~/repos/my-app

# As MCP — agents call tools directly
railguey_deploy   workspace: "/path" service: "web"
railguey_status   workspace: "/path"
railguey_doctor   workspace: "/path"

17 GraphQL-based tools. Project-scoped tokens — no railway login needed. Same token works in local dev, CI/CD, and AI agents.