Qalibr docs — your guide to bug intake intelligence. Start a free trial →
Core conceptsCalibration
Core concepts

Calibration

“Calibration” shows up in two places in Qalibr. Both come back to the same idea: the scoring tunes itself to your team, and it’s honest about how settled it is.

Scores that settle in

When you’re just starting, there aren’t many routed bugs to learn from, so Signal scores carry a [ PROVISIONAL ] tag — scores calibrate to your inbox as you route more bugs. As you approve and route more reports, Qalibr has more to anchor on, the scores settle, and the tag quietly disappears. Calibrated is the silent, steady state.

Honest markers

Calibration also shows up on individual severities. A ~ marker means Qalibr is hedging — it’s offering a starting point rather than a firm call. A marker means a deterministic rule raised the severity for a critical surface. Together they tell you, at a glance, which calls to trust and which to double-check. See Severity.

The Calibration view

The dashboard has a Calibration view — the bugs sitting in Review, waiting for your decision. Every time you approve or reject one, you make the call the scoring learns from. Working that queue is how you calibrate Qalibr to your team’s bar for what matters.

There’s nothing to configure here. Calibration happens as a side effect of you doing the normal thing — triaging your inbox.