# Wazuh Rule Tuning & False Positive Reduction (lab)

> A utility's SOC has stopped reading its alert queue: one rule produces nearly all of it. Three tuning proposals are waiting for sign-off and every one of them genuinely reduces volume.

Canonical: https://codelivly.com/labs/wazuh-rule-tuning

| | |
|---|---|
| Type | Investigation |
| Difficulty | Intermediate |
| Points | 350 |
| Session length | 45 min |
| Completions | 2 |

A utility's SOC has stopped reading its alert queue: one rule produces nearly all of it. Three tuning proposals are waiting for sign-off and every one of them genuinely reduces volume. Two of them would also have hidden a real intrusion. Work out which is which, and be specific about what the others would have blinded the SOC to.

## Objective

Learn the measurement that separates detection engineering from alert deletion. You will quantify what a rule change removes, then test the same change against a regression set of alerts already confirmed as real incidents, and see why the proposal with the best volume figure can be the most dangerous one on the table.

## Tasks

1. Establish the baseline
2. Read all three proposals before evaluating any
3. Measure what each proposal removes
4. Replay every proposal against the regression set
5. Explain the blind spots, not just the numbers

## Related

- [All labs](https://codelivly.com/practice?type=lab)
