Built a terminal network incident tool that now exports frozen bundles with pcap + context
I built NetWatch because debugging network issues from a terminal usually turns into a pile of tools: one for interface throughput, one for process/socket visibility, one for pings, one for tracero...

Source: DEV Community
I built NetWatch because debugging network issues from a terminal usually turns into a pile of tools: one for interface throughput, one for process/socket visibility, one for pings, one for traceroute, and then Wireshark/tshark when I finally need packets. NetWatch tries to collapse that into one terminal UI. It shows live interface rates, active connections with process name + PID, gateway/DNS health, packet capture with decode/filtering, topology/traceroute, a connection timeline, and per-process bandwidth. The new thing in v0.9.0 is a rolling "flight recorder". You can arm a 5-minute window, let it keep rotating, then freeze/export when the issue actually happens. The export bundle includes a pcap plus connection snapshots, health samples, bandwidth context, DNS analytics, alert history, and a summary. The goal is to preserve transient incidents that usually disappear before you can inspect them. It runs locally and doesn't require a backend or service. netwatch gives interface/conn