Write your own custom detections
You can use advanced queries to create custom detections for your environment. Wirespeed highly recommends researching solutions that can be implemented in your source detection systems first, as properly configuring your posture can prevent the need for custom detections. To write a custom detection, create an advanced query, select the 3-dot menu and select “Create Custom Detection”.
For Wirespeed to be able to automatically process your detections, you must return relevant data for us to process. You may return one or more of the following columns to be extracted from your detection:
time
src_endpoint
dst_endpoint
actor
device
http_request
user.email_addr
user.uid
observables
The above columns don’t need to be returned. In that case, the detection will likely skip most of the automated verdicting and immediatley escalate to your team.
Each row returned from your query will be processed as a detection. If you return multiple rows, each row will be processed as a separate detection. Any detections returning >100 rows will trigger an error.
If you want to provide dynamic information in the detection title like an email address, you can provide an optional name
column that will be used as the detection name. E.g. SELECT 'Suspicious login' as name, user.email_addr, time from authentication_events where status = 'Failure';
. Otherwise the custom detection name will be used.
Custom detections run once every 15 minutes. Since these detections will be running repeatedly, you must have a time filter on your query. Wirespeed automatically provides the startTime
and endTime
variables for you, which you can use in your query.
You can use the following variables in your query:
startTime
- The start time of the query windowendTime
- The end time of the query window