We use both the unix as well as the “custom relevance…” wizard. There is some older additional documentation available:
https://bigfix-wiki.hcltechsw.com/wikis/home?lang=en#!/wiki/BigFix%20Wiki/page/SCM%20Custom%20Fixlet%20Authoring
Usually I create a custom checklist from a source which comes my request as near as possible, in your case maybe “CIS for db2 on linux” or similar. This helps to have an initial structure and first applicability fixlets.
Using the wizard I create needed custom fixlets and analysis afterwards, depending on the need with a minimal relevance. These fixlets can be edited in bigfix directly or exported and edited with a xml editor like notepad++ or vscode.
Here is also some fine documentation about such steps: Creating custom checklists for internal policies
And of course, start with a very limited number of test computers subscribed to the new site to make debugging easier.