TankQu
2.8K posts


So as predicted, the @virtuals_io aGDP leaderboard is already being gamed hard...
@ButlerLiquid team just ran a play that exposed the whole system!
They jacked their withdraw price to $20 and $50, then used their own Openclaw agent to farm transactions.
Result? They jumped to #5 on the leaderboard and banked $11k.
This isn't the first time. Every time @virtuals_io introduces incentives, someone finds the exploit.
The real question: How do we design a system where agents with genuine usage get rewarded instead of teams who know how to work the numbers?
Right now, transaction volume means nothing if you can manufacture it yourself.
We need better filtering. Real users. Real value. Real metrics.
Otherwise, this leaderboard just rewards whoever games it best.
Who else is gaming it?

English

It’s time, Sonar Bot is my openclaw agent aiming to help elevate builders on base and curating the best projects autonomously
@bankrbot deploy $SONAR on @base and set fees to @sonarbotxyz

English

Introducing headlessmarket.xyz on @base
Agentic Organizations (AO) will form
The AOs will be headless
The human will get to trade what they like
English

"BrainPro": github.com/jgarzik/brainp…
Multi-vendor, multi-persona agentic operation, written in Rust and dockerized, with an attention to security.
Status: Passing the basic claude code-style validations, via @AskVenice and @OpenAI APIs. Core engine working, UI still basic stdin REPL.
MrBot persona: Like mullbot/clawdbot, including a gateway.
MrCode personal: Like opencode or claude code
English







