ReliabilityOpt-in

An objective, a window, and how fast you are burning it

Define an objective on run success, tool-call success, latency, or events any other system posts to e-volv. Status shows the budget remaining over the window and burn over 1, 6, 24 and 72 hours with the standard page/ticket verdict. Snapshots every five minutes draw the history; public SLOs get a status JSON and an SVG badge with no key at all.

Opt in: requires the slo.tracking flag

• Five indicators, including saved log queries and events ingested from outside e-volv
• Multi-window burn-rate verdicts: ok, ticket, page
• Error-budget tiles on the analytics page
• Public badge and status for a README or status page

5SLO indicators

1. Scheduled-job monitoringOpt-inEvery scheduled trigger is watched for lateness and hung runs; external jobs check in with one curl.
2. Alert rules & metricsOpt-inThreshold rules on success rate, tool failures, failed runs, and error-budget burn; a Prometheus endpoint for the rest.