AI Rules Moderation — 14-day trial, then $25/year

AI Rules Moderation — 14-day trial, then $25/year 1.0.8

No permission to download
Overview FAQ Updates (9) History Discussion
zeeb0t

zeeb0t

Well-known member
Licensed customer
zeeb0t submitted a new resource:

AI Rules Moderation - Plain-language AI moderation rules for XenForo, with flagged content sent to the approval queue.

AI Rules Moderation helps XenForo forum owners enforce their own plain-language forum rules using AI, without making the forum server do the expensive processing.

The add-on watches supported XenForo content changes, sends eligible visible content to our externally hosted moderation service, and moves clearly flagged content into XenForo's approval queue for human review.

It is designed to assist moderators, not replace them.

What it does​

  • Lets you write...
Click to expand...

Read more about this resource...
 
zeeb0t updated AI Rules Moderation with a new update entry:

Release: v1.0.1

  • Added Admin CP helper text showing the current XenForo Board title and Board meta description as suggested values for forum name and forum about.
  • Changed AI Rules Moderation to use a dedicated least-privilege API scope instead of a super-user/all-scopes API key.
  • Added an Options menu link to the add-on controls menu, linking directly to the AI Rules Moderation setup page.
  • Updated the add-on icon to use a custom image asset instead of the default Font Awesome shield.
Click to expand...

Read the rest of this update entry...
 
Does it use a credit for every post it checks or does it use a credit when it takes an action on a post. Nice looking mod, it is clean. I wouldn't want it to go through over a million posts on my forum, Does it ignore older posts before the add on is installed?
 
greenchicken said:
Does it use a credit for every post it checks or does it use a credit when it takes an action on a post. Nice looking mod, it is clean. I wouldn't want it to go through over a million posts on my forum, Does it ignore older posts before the add on is installed?
Click to expand...

Thanks! A moderation credit is used when the hosted service checks eligible content, not only when it takes action. In other words, if a new or updated post is sent for AI review, that review consumes one credit whether the post is allowed through or moved to the approval queue.

It does not crawl or bulk-check old forum content when the add-on is installed, so it will not suddenly run through a million existing posts. The add-on is event-based and is intended to check supported content when it is newly created or updated after the integration is active. Older posts are ignored unless they are edited/updated in a way that triggers review.

We also maintain a short-lived moderation cache to help avoid unnecessary repeat credit usage. If the same content has already been checked within the last 2 weeks and has not actually materially changed, or it is changed back to a version that was already checked, it is skipped rather than being sent for another moderation check.

We’ve also just added a 14-day trial, which includes 100 test moderation checks with no billing during the trial period. That should make it easier to test on a live forum and confirm the behaviour before committing. For larger/high-volume forums that need a more representative trial allowance, feel free to contact us and we can discuss that.

The goal is to keep it lightweight and predictable: new/changed eligible content gets checked, clearly flagged content goes to the normal XenForo approval queue for staff review, repeated unchanged checks are avoided where possible, and historical content is not backfilled automatically.
 
zeeb0t updated AI Rules Moderation with a new update entry:

Release: v1.0.3

  • Added an Admin CP audit log for AI Rules Moderation outcomes.
  • Logs both moderated content and checked content where no action was taken.
  • Added support for log_only callback results without changing content state.
  • Added audit log filtering, sorting, pagination, and CSV export.
  • Stores additional moderation metadata, including content date, logged date, outcome/action, status, and reason.
  • Updated the approval queue reason display to use the new audit log data.
  • Added...
Click to expand...

Read the rest of this update entry...
 
is it possible for it to work on one specific node where all the problem children hang out such as discussing politics forum?

This is the only place I get dinged by google adsense as I don't really want or need full forum moderation
 
greenchicken said:
is it possible for it to work on one specific node where all the problem children hang out such as discussing politics forum?

This is the only place I get dinged by google adsense as I don't really want or need full forum moderation
Click to expand...

Nice idea. I will add an update in the next few days that allows you to select one or many nodes to run on, and ignore the rest. I will tag you in the update when it's available.
 
zeeb0t said:
Nice idea. I will add an update in the next few days that allows you to select one or many nodes to run on, and ignore the rest. I will tag you in the update when it's available.
Click to expand...
please do, I am very interested in your mod but I don't want AI to interfere anywhere I don't want it to , maybe install usergroup permissions as well to disable checking on specified usergroups
 
greenchicken said:
please do, I am very interested in your mod but I don't want AI to interfere anywhere I don't want it to , maybe install usergroup permissions as well to disable checking on specified usergroups
Click to expand...

Another great idea. I will also implement a similar option for user groups! Thank you.
 
zeeb0t updated AI Rules Moderation with a new update entry:

Release: 1.0.6

For XenForo <v2.3.8
  • Older XenForo versions that do not safely support webhook criteria continue using the previous broad webhook behaviour (all nodes, all users, all content types).
For XenForo >=v2.3.8
  • Added new AI moderation scope settings for nodes, user groups, and supported content types within the existing AI rules/settings page.
  • Added XenForo-native webhook criteria integration for scoped dispatch, preventing non-matching content from being...
Click to expand...

Read the rest of this update entry...
 
greenchicken said:
please do, I am very interested in your mod but I don't want AI to interfere anywhere I don't want it to , maybe install usergroup permissions as well to disable checking on specified usergroups
Click to expand...

The latest 1.0.6 release now supports your suggested features around node/group/content type selections. Please note the required minimum version for XenForo is v2.3.8. Any version v2.3.0 <-> 2.3.7 will work, but without this new addition. This new addition requires version v2.3.8 or newer.
 
zeeb0t updated AI Rules Moderation with a new update entry:

Release: 1.0.7

  • Added XenForo 2.1 and 2.2 support for sites running PHP 7.2+.
  • Added a compatibility webhook system for older XenForo versions that do not include native XF 2.3 webhooks.
  • Preserved the existing XenForo 2.3+ native webhook implementation unchanged.
  • Improved moderation dispatch reliability with queued delivery, duplicate protection, stale-content checks, and safer version-aware class loading.
  • Maintained existing approval queue and moderation callback...
Click to expand...

Read the rest of this update entry...
 
This is a cool addon.
Could it be used for spam control? our forum has one topic, so a good rule would be, are the user's first posts having anything to do with that.

We have our own local AI hardware running over the OpenAI protocol. Could we use it in a future version?
 
ES Dev Team said:
This is a cool addon.
Could it be used for spam control? our forum has one topic, so a good rule would be, are the user's first posts having anything to do with that.

We have our own local AI hardware running over the OpenAI protocol. Could we use it in a future version?
Click to expand...

I can certainly add an option to provide the first post content as context for the new message - and a supporting rule would enforce on-topic with original post etc.

And yes, in a future version there will be a BYO key and even BYO openai api compatible endpoint. Those versions will require a single payment for the addon rather than the current pay as you go.
 
You must log in or register to reply here.
Back
Top Bottom