菜单
Check Railway services (api/web) for errors, view logs, and fix deployment issues. Use when deployments fail, services are unhealthy, or you need to investigate production errors.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Research and save a book, video, blog post, article, paper, podcast, link, or other resource into memory. Use when the user says "/save", "save this", "remember this", "store this resource", or asks Claude to quickly research something and log it into `~/Dropbox/memory/resources/`.
Use when creating a pull request, opening a PR, or when work on a branch is ready for review
Scan an entire repository and produce a long-term improvement plan. Use when the user says "improvement plan", "/improvement-plan", "roadmap", "what should we improve", "long-term plan", "tech debt audit", "codebase review", or asks to analyze the project for future improvements, refactoring opportunities, or strategic technical direction.
Ask clarifying questions when a task is ambiguous or underspecified before implementing. Identifies missing objectives, acceptance criteria, scope, constraints, and environment details, then presents concise numbered questions with multiple-choice options and sensible defaults. Do not use automatically, only when invoked explicitly
Systematic workflow for filling out PR templates with complete documentation and analysis
Simplifies and refines code for clarity, consistency, and maintainability while preserving all functionality. Focuses on recently modified code unless instructed otherwise.
| name | railway-debug |
| description | Check Railway services (api/web) for errors, view logs, and fix deployment issues. Use when deployments fail, services are unhealthy, or you need to investigate production errors. |
| allowed-tools | Bash(railway *), Read, Grep, Glob |
| argument-hint | [service] |
Check Railway services for errors and fix them.
Usage: /railway-debug or /railway-debug [service] where service is api, web, or all.
First, check which services are healthy or failed:
railway service status --all
For each failed or problematic service, check logs in this order:
Build logs (compilation/bundling issues):
railway logs --service <service> --build --lines 100 --latest
Deployment logs (runtime errors):
railway logs --service <service> --lines 100 --latest
Error-specific logs:
railway logs --service <service> --lines 50 --filter "@level:error" --latest
| Error Pattern | Likely Cause | Fix |
|---|---|---|
Module not found | Missing dependency | Check web/package.json, run npm install |
TypeScript error | Type issue | Run npx tsc --noEmit in web/ |
Build failed | Vite config issue | Check web/vite.config.ts |
CORS error | API URL mismatch | Check VITE_API_URL env var in Railway |
| Error Pattern | Likely Cause | Fix |
|---|---|---|
Migrations pending | DB out of sync | Run railway run bin/rails db:migrate |
Bundler error | Gem issues | Check api/Gemfile, ensure Ruby version matches |
SECRET_KEY_BASE | Missing env var | Set in Railway variables |
Database connection | Postgres URL | Check DATABASE_URL is set from Postgres service |
After identifying the issue:
cd web && npm run buildcd api && bundle exec rspecAfter fixes are verified:
railway redeploy --service <service>
| Service | Directory | Description |
|---|---|---|
| api | api/ | Rails backend |
| web | web/ | Vite/React frontend |
| Postgres | - | Database (check connection issues) |
DATABASE_URL - Should reference the Postgres serviceRAILS_ENV - Should be productionSECRET_KEY_BASE - Must be set for productionRAILS_MASTER_KEY - If using credentialsVITE_API_URL - API endpoint URLNODE_ENV - Should be production for builds