← All Tools

Sigma Rule Builder

Build a valid Sigma detection rule from a form, preview the YAML live, and export best-effort Splunk, QRadar and Sentinel queries.

Everything runs in your browser. Your rule is never uploaded or stored on our servers. Targets SigmaHQ sigma-specification v2.0.0.

Metadata

Generated with your browser's secure RNG.
MITRE ATT&CK style, e.g. attack.t1059.001.

Log source (at least one of category / product / service)

Detection

A rule has one or more named search identifiers (selections / filters) plus a condition that combines them. Each field inside a selection is AND-ed; multiple values are OR-ed (or AND-ed with "all values").

Defined identifiers: (none yet). Grammar: and or not, parentheses, 1 of selection_*, all of them.

Known false positives

YAML preview



                    

SIEM export (best-effort, not pysigma)


                        

Local library (this browser only)

How to use

  1. Fill in a title and a log source (e.g. category process_creation, product windows).
  2. Add one or more selections. In each, pick a field (e.g. Image), a modifier (e.g. endswith) and one value per line.
  3. Write a condition that references your selection names, e.g. selection_img and selection_cmd and not filter_sccm. Use Suggest from selections for a starting point.
  4. Read the validation line, copy the YAML, and check the SIEM export tabs.

What the SIEM export does — and does not — do

The export is a best-effort translation, not pySigma. It faithfully transpiles the string operators (equals, contains, startswith, endswith), their list semantics (OR, or AND with all values), numeric comparisons, exists, and the full condition logic (and/or/not, parentheses, N of/all of them). Field names are passed through verbatim — map them to your schema (CIM, ECS, etc.) yourself. Modifiers with no faithful inline equivalent in a given backend (for example re on Splunk/QRadar, cidr on Splunk, base64, base64offset, fieldref) are never silently approximated: the field is omitted and listed in the notes so you can translate it by hand. For production pipelines, generate backend queries with pySigma.

Worked example

Press Try an example for a realistic process_creation rule (a PowerShell download cradle) with two selections and a negated filter, matching the SigmaHQ rules repository style.

References