展示HN：OneUptime – 一款开源的可观测性工具，利用人工智能自动修复事件
嘿，HN，
我这几年一直在开发 OneUptime。这是一个完整的开源可观测性平台，能够替代 Pingdom、StatusPage.io、PagerDuty、Datadog 和 Sentry——所有功能合而为一。
它的不同之处在于：我们正在朝着自主事件解决的方向发展。我们的 AI 代理不仅仅是在凌晨 3 点提醒你，而是能够检测问题、生成代码修复并打开拉取请求。你醒来时看到的是修复，而不是警报。
功能包括：
- 来自多个全球位置的正常运行时间监控
- 公共/私人状态页面（免费，无限制）
- 带时间线的事件管理
- 值班调度与升级策略
- 日志、追踪、指标、错误跟踪
- 原生支持 OpenTelemetry
可以使用 Docker/Kubernetes 自行托管，或使用我们的云服务。采用 Apache 2.0 许可证。
代码库： [https://github.com/OneUptime/oneuptime](https://github.com/OneUptime/oneuptime)
演示： [https://oneuptime.com](https://oneuptime.com)
希望能得到反馈，特别是关于 AI 自动修复功能的反馈。什么样的条件会让你信任 AI 来修复生产问题？
查看原文
Hey HN,<p>I've been building OneUptime for the past few years. It's a complete open-source observability platform that replaces Pingdom, StatusPage.io, PagerDuty, Datadog, and Sentry – all in one.<p>What makes it different: We're building toward autonomous incident resolution. Instead of just alerting you at 3am, the AI agent detects the issue, generates a code fix, and opens a PR. You wake up to fixes, not pages.<p>Features:<p>• Uptime monitoring from multiple global locations<p>• Public/private status pages (free, unlimited)<p>• Incident management with timelines<p>• On-call scheduling with escalation policies<p>• Logs, traces, metrics, error tracking<p>• OpenTelemetry native<p>Self-host with Docker/Kubernetes or use our cloud. Apache 2.0 licensed.<p>Repo: <a href="https://github.com/OneUptime/oneuptime" rel="nofollow">https://github.com/OneUptime/oneuptime</a>
Demo: <a href="https://oneuptime.com" rel="nofollow">https://oneuptime.com</a><p>Would love feedback, especially on the AI auto-fix feature. What would make you trust an AI to fix production issues?