今天我们推出由 @ycombinator 支持的 Hyperprobe：一个在生产环境中实时调试故障代码、把事故追到已确认根因的值班 Agent。
今天我们推出由 @ycombinator 支持的 Hyperprobe：一个在生产环境中实时调试故障代码、把事故追到已确认根因的值班 Agent。AI 如今写下大部分代码，但代码在生产环境出错时，能解释问题的值往往从未被记录；工程师只能补日志、重新部署，再等待故障重现。HyperProbe 跳过这轮循环：它在运行中服务的确切代码行放置只读探针，捕获故障请求命中的变量、调用栈和日志中没有的证据。coding agent 可通过 MCP 端到端运行它，并返回有真实数据支撑的根因。作者举例称，9 月印度一家上市金融科技公司因失败交易被标记为成功而赔付 500 万美元；团队此前花了三年研究如何在不破坏线上服务的前提下提取运行时状态。
查看英文原文
Today we're launching Hyperprobe backed by @ycombinator: an on-call agent that debugs failing code live in production and takes an incident to a confirmed root cause. AI writes most of our code now, but when it breaks in prod the value that explains it was never logged, so engineers add a log, redeploy, and wait for the bug to fire again. HyperProbe skips that loop. It drops a read-only probe on the exact line in your running service and captures the state a failing request hits: variables, call stack, the evidence logs never had. Your coding agent runs it end to end over MCP and hands back a root cause backed by real data. In September a public fintech in India paid out $5 million after failed transactions were marked successful. Same pattern: code ran clean, wrote the wrong state, nothing in the logs showed it. We spent three years pulling runtime state out of live services without breaking them. Debugging code you didn't write at 2 AM? Try it: http://docs.hyperprobe.co/quickstart
Today we're launching Hyperprobe backed by @ycombinator: an on-call agent that debugs failing code live in production and takes an incident to a confirmed root cause. AI writes most of our code now, but when it breaks in prod the value that explains it was never logged, so engineers add a log, redeploy, and wait for the bug to fire again. HyperProbe skips that loop. It drops a read-only probe on the exact line in your running service and captures the state a failing request hits: variables, call stack, the evidence logs never had. Your coding agent runs it end to end over MCP and hands back a root cause backed by real data. In September a public fintech in India paid out $5 million after failed transactions were marked successful. Same pattern: code ran clean, wrote the wrong state, nothing in the logs showed it. We spent three years pulling runtime state out of live services without breaking them. Debugging code you didn't write at 2 AM? Try it: http://docs.hyperprobe.co/quickstart
Today we're launching Hyperprobe backed by @ycombinator: an on-call agent that debugs failing code live in production and takes an incident to a confirmed root cause. AI writes most of our code now, but when it breaks in prod the value that explains it was never logged, so engineers add a log, redeploy, and wait for the bug to fire again. HyperProbe skips that loop. It drops a read-only probe on the exact line in your running service and captures the state a failing request hits: variables, call stack, the evidence logs never had. Your coding agent runs it end to end over MCP and hands back a root cause backed by real data. In September a public fintech in India paid out $5 million after failed transactions were marked successful. Same pattern: code ran clean, wrote the wrong state, nothing in the logs showed it. We spent three years pulling runtime state out of live services without breaking them. Debugging code you didn't write at 2 AM? Try it: http://docs.hyperprobe.co/quickstart
- 来源
- X @ycombinator · x-ycombinator
- 证据角色
- x_project_network
- 时间状态
- known
- 用户判决
- 未判