We need to decide on the direction we will take to enable community administrators to configure Automoderator.
We may consider using [[ https://www.mediawiki.org/wiki/Community_configuration_2.0 | Community Configuration ]], which is under active development by the Growth team. They have offered engineering support in Q3 if we go this route.
The kinds of things we expect communities to need to configure include:
* Mode: Log-only or Reverting (toggle)
* Caution level (number or selection)
* Users should also be able to fine-tune the threshold by changing the number directly even if this is a selection in the UI.
* Whether the bot should use the bot flag or not
* What edit summary should be used when reverting
* A link to a MediaWiki page to define the talk page message
* How frequently the tool can revert the same editor on a page
**User stories**
* As an administrator, I want to change Automoderator’s configuration, so that I can fine-tune the settings to reflect what behaviour my community is looking for.
* As a patroller, I want to see what changes have been made to Automoderator's configuration, so that I can review how other editors changed the tool's behaviour in the past and have transparency into the tool.
* As a patroller, I want to know when changes are made to Automoderator, so that I can understand changes in its behaviour.
* As a patroller, I want to know what Automoderator’s current configuration is, so that I can understand how it will behave.
**Technical**
There are some notes in T345092, but we want to further think about:
- Is Community Configuration a good fit?
- How will we access these configurations from the edit reverting component? (T349295)
-
**Design goals** - T349238
We want changes to Automoderator's configuration to be transparent, easily visible, and for there to be a clear log.
We may want to build novel illustrations or interactive elements into the configuration interface, for example to help users understand the impact of changing the caution level.