Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

High-risk tag autoreport and moderation queue #147

Open
TheSmilingPony opened this issue Nov 18, 2021 · 0 comments
Open

High-risk tag autoreport and moderation queue #147

TheSmilingPony opened this issue Nov 18, 2021 · 0 comments
Labels
enhancement New feature or request

Comments

@TheSmilingPony
Copy link

Problem:
Some very special people like to mass vandalize tags, usually dozens or hundreds at a time. Often these go unnoticed for hours, or even days, until some unfortunate soul does a in innocent search for something "safe" and is met with definitely "unsafe" images.

Describe the solution you'd like
A hitman.

Describe alternatives you've considered
Staff can flag certain tags as "High Risk" or "Monitor Changes". Whenever a tag with this flag is removed or added from an image, a report is generated in a new, separate moderation queue, "T". A separate queue is desirable in order to highlight the problem (as it should be dealt with in a definite manner quickly) and not affect the normal Report queue (often filled with low priority or claimed tickets).

Reports can be generated per image affected by the monitored tag, or one single report per monitored tag per IP/user (as staff can then track tag changes made by that IP/user).

Staff can mass mark / tick reports as claimed / closed, as tag vandals often hit many multiples of images at once.

Additional context
None that wouldn't get this repo deleted.

@TheSmilingPony TheSmilingPony added the enhancement New feature or request label Nov 18, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

1 participant