C O M P Ξ Z
@compez.eth
Updated... β A few steps to the final backend... This version does several things at same time. 1) Process all users you follow. 2) Process all users provided by WC Label 3) Parse all spammed users 4) Match spam users with users you follow.
7 replies
0 recast
14 reactions
Pichi πͺππΉπ© π‘πΈ
@pichi
So cool. Now how to match FID to username so I can investigate them?
1 reply
0 recast
3 reactions
C O M P Ξ Z
@compez.eth
Yeah! We can generate them along with fids. It may take a little longer to process, but it is possible. After the main logic in processing, I solve transparency from a ux perspective.
1 reply
0 recast
2 reactions
Pichi πͺππΉπ© π‘πΈ
@pichi
This is so awesome. If Iβm following people, I donβt think they are spammy. This is a great way for the average person to help the Warpcast team identity false positives. Maybe they really did go off the rails and start behaving differently but maybe they got flagged by mistake. Imagine how many folks we can help with this! Or we learn we are following bad accounts and unfollow them and block if they are bots pretending to be humans!
1 reply
0 recast
1 reaction
C O M P Ξ Z
@compez.eth
I agree! It was interesting In the first test out of 5 spams I removed 3 of them! But the other two are not spam but interact with spammers! We can make this a part of the complete analytics tool and provide a transparent analyzed profile page for users. In subsequent attempts, we will achieve more useful and accurate results with more details.
1 reply
0 recast
1 reaction
Pichi πͺππΉπ© π‘πΈ
@pichi
YES! Iβm going to go manually look up my people and report back!
1 reply
0 recast
1 reaction
Pichi πͺππΉπ© π‘πΈ
@pichi
One is an AI agent I created @kyotoguide via Janβs new AI agent. @elxlee looks like the agent got a spam label already and Iβve only called on it a few times! Might need to work with the Warpcast team to get them labeled as good bots!
3 replies
0 recast
2 reactions