home / skills / einverne / dotfiles / deploy-staging

deploy-staging skill

needs review

This skill deploys the current branch to staging after pre-checks, builds, runs deployment, and validates health, reporting results and offering rollback.

npx playbooks add skill einverne/dotfiles --skill deploy-staging

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

1.3 KB

---
name: deploy-staging
description: 将当前分支部署到测试环境。当用户要求部署、发布到测试或在 staging 环境测试时使用
---

# 部署到测试环境

## 前置条件检查

执行部署前，必须验证：

1. 当前分支所有变更已提交
   - 运行 `git status` 检查
   - 如有未提交内容，提示用户先提交

2. 所有测试通过
   - 运行 `npm test`
   - 如测试失败，停止部署并报告失败原因

3. 代码已推送到远程仓库
   - 运行 `git push`
   - 确保远程分支是最新状态

## 部署流程

1. 构建生产版本
   ```bash
   npm run build:staging
   ```

2. 运行部署脚本
   ```bash
   ./scripts/deploy-staging.sh
   ```

3. 等待部署完成
   - 显示部署进度
   - 捕获并显示错误信息

4. 验证部署
   ```bash
   curl -f https://staging.example.com/health
   ```

5. 报告部署结果
   - 成功：提供访问链接
   - 失败：显示错误日志和回滚建议

## 错误处理

如果部署失败：
1. 保存错误日志到 `.logs/deploy-staging-{timestamp}.log`
2. 建议用户检查日志
3. 提供回滚命令：`./scripts/rollback-staging.sh`

## 部署后检查

部署成功后，提醒用户验证：
- 关键功能是否正常
- 数据库迁移是否成功
- 静态资源是否加载正常

Overview

This skill deploys the current Git branch to a staging environment with pre-checks, build, deploy, verification, and post-deploy guidance. It automates the common steps required to push a branch to staging and surfaces failures, logs, and rollback options. Use it when you want a repeatable, safe staging release process.

How this skill works

The skill first validates repository state: ensures all changes are committed, tests pass, and commits are pushed to the remote. It runs a staging build, executes the staging deployment script, streams progress and errors, then performs a health check against the staging URL. On failure it saves logs, shows error details, and suggests a rollback command.

When to use it

You want to deploy the current branch to the staging environment for testing.
Before requesting QA or stakeholder review on staging.
When preparing a pre-production verification after feature work.
After merging feature work but before production release.

Best practices

Commit or stash all changes before running the deploy to avoid partial deploys.
Run the full test suite locally (npm test) and fix failures before deploying.
Push your branch to the remote so the staging environment builds from the expected ref.
Monitor the deployment output and run the health check (curl -f https://staging.example.com/health) after completion.
Save and inspect the timestamped log file if deployment fails, then use the provided rollback script.

Example use cases

Developer finished a feature branch and needs it available on staging for QA.
A bug fix that must be validated in an environment mirroring production.
Release candidate verification before scheduling a production deploy.
Automated CI job triggers a manual staging deployment when preconditions are met.

FAQ

What happens if tests fail before deployment?

Deployment stops and the failing test output is reported. Fix tests locally and re-run the deploy.

How do I roll back a failed staging deploy?

Inspect the saved log file (.logs/deploy-staging-{timestamp}.log) for root cause, then run the rollback script: ./scripts/rollback-staging.sh