The only tool you need to know what is happening and how to fix it.
Go alertingexception-trackerissue-trackinglogsmetricsmonitoringobservability-platformopen-telemetryotelsession-replaytelemetrytraces
Ground Rules (All Flows)
- Reads are safe: any
list/show/querysubcommand may run freely; they never mutate server state. - This CLI is read only:
tracewayhas no mutating commands, and its login can only read. Changes (archiving issues, acknowledging pages, snoozing alerts, editing post-mortems) need the separatetraceway-writeCLI and thetraceway-writeskill. "Look at this error" means read it, never change it. - Output: piped output defaults to JSON (table on a TTY). Prefer JSON +
jq, and--fields a,b,cto trim responses. Keep--page-sizeat 10 to 20 for triage. - Time windows: always bound queries, default
--since 1hfor "now" questions,--since 24hotherwise.--sinceacceptss,m,h, lowercaseNd(no1w, no7d2h). Absolute windows via--from/--to(RFC3339). - Exit codes: 0 ok, 1 generic/API, 2 usage, 3 connection, 4 auth, 5 not found, 6 rate limited, 7 server 5xx. Errors emit
{"error":"<stable_id>","message":"...","hint":"...","exit_code":N}on stderr; branch on theerrorfield. - On exit code 4 (auth), do not run
traceway loginyourself; switch to the Login flow and let the user enter credentials. - Never guess a fix: if the telemetry does not explain the failure, do not change behaviour on a hunch. Add the instrumentation that would explain it next time, and say what is still unknown. See "When the evidence is not enough" in the Debug flow.