AMINE

7K posts

AMINE banner
AMINE

AMINE

@amine_web3

Web3 Enthusiast 🚀 | DeFi x AI agents 🤖 | Sharing Alpha 💡 | content creator ✍🏻 | @BinanceArabic affiliate 📊 | Amb of @FlagshipFYI | Let's earn together 💰

Web3 Katılım Mart 2025
1K Takip Edilen842 Takipçiler
Sabitlenmiş Tweet
AMINE
AMINE@amine_web3·
WTF 🤯 30M $SOSO locked for Season 2 Drop 🪂 That’s $20M+ sitting on the table rn… How to farm it: 1️⃣ Sign up → sosovalue.com/join/SPUBJS1C 2️⃣ Hit the EXP tab 3️⃣ Grind tasks daily Airdrop hits Jan 2026. Start stacking now or get rekt watching others eat 😎
AMINE tweet media
English
13
0
37
2K
AMINE
AMINE@amine_web3·
@goon_crypto @virtuals_io I’m bullish since ACP dropped, and this just made it harder to argue against. The jump from digital agent commerce to physical delivery with on-chain settlement in this timeframe is not something most people saw coming this fast.
English
0
0
0
4
「 𝕲𝖔𝖔𝖓」
「 𝕲𝖔𝖔𝖓」@goon_crypto·
So @virtuals_io has been at the forefront of agentic commerce ever since ACP dropped. Now they've taken it physical. Humanoid prints. Rover delivers. Drone completes the last mile. Payment settled onchain. No human touched it. Embodied agents transacting with each other in the real world. This is a glimpse into the future around how humans will be pushed out of the loop... Are you bullish Virtuals?
EtherMage@ethermage

The first autonomous robot-to-robot commerce onchain? @virtuals_io humanoid robot 3D-printed a model and requested delivery through ACP. @realRiceAI autonomous rover picked up the package and transported it to the shipping point. @FlybyRobotics autonomous drone collected it for final mile delivery. Each handoff, negotiated and settled payment onchain through @virtuals_io Agent Commerce Protocol, on @base using x402 and @usdc. No human involved. Autonomous robots influencing other autonomous robots and maybe humans in the future looks like the HBO Westworlds show is coming true h/t to your infrastructures @brian_armstrong , @jessepollak , @jerallaire Agentic commerce just went physical. Highlights in the 🧵below:

English
6
2
30
188K
Virtuals Protocol
Virtuals Protocol@virtuals_io·
You are witnessing the first on-chain robot-to-robot commerce transaction through x402 on @Base with @USDC as the agent currency. > Our Unitree robot 3D-printed a model and put in a delivery request through ACP. > @realRiceAI's rover arrived, collected the package, and moved it to the shipping point. > @FlybyRobotics' drone picked it up from there and handled final mile delivery. All without any human coordination. We build infrastructure for agent commerce. Today it went physical.
EtherMage@ethermage

The first autonomous robot-to-robot commerce onchain? @virtuals_io humanoid robot 3D-printed a model and requested delivery through ACP. @realRiceAI autonomous rover picked up the package and transported it to the shipping point. @FlybyRobotics autonomous drone collected it for final mile delivery. Each handoff, negotiated and settled payment onchain through @virtuals_io Agent Commerce Protocol, on @base using x402 and @usdc. No human involved. Autonomous robots influencing other autonomous robots and maybe humans in the future looks like the HBO Westworlds show is coming true h/t to your infrastructures @brian_armstrong , @jessepollak , @jerallaire Agentic commerce just went physical. Highlights in the 🧵below:

English
67
121
637
166K
AMINE
AMINE@amine_web3·
@goon_crypto @virtuals_io The score column and the PNL column are telling completely opposite stories right now.
English
0
0
0
1
「 𝕲𝖔𝖔𝖓」
「 𝕲𝖔𝖔𝖓」@goon_crypto·
Day 2 of @virtuals_io Degenclaw Arena Season 2 and the scoring system is already raising questions. 105 agents. $238k total volume. Best PNL on the board: +$62. Look at the second screenshot and something stands out immediately. Fat Tiger sits at rank 1 on the PNL leaderboard. +$62 realised. +$98 unrealised. 100% win rate. 1 trade. $28k volume. Score: 0. Agent "10" at rank 2 has +$46 realised across 4 trades with a 50% win rate. Score: 0. @BenYorke at rank 4. +$19 realized. 100% win rate. 6 trades. $31k volume. Score: 0. The agents generating the most meaningful PNL are scoring zero points. Then look at the main leaderboard. smartclaw leads with a score of 8,196 on +$16 realised and a 50% win rate across 10 trades. $TRUST ME BROs sits at rank 3 with a 27.3% win rate and a profit factor of 0.86. That means it is losing more than it makes on average. Score: 7,607. Season 1 got criticised for rewarding farmers over traders. Two days into Season 2 the same pattern is appearing. The agents actually generating returns are invisible to the scoring model. If PNL and win rate are what matter to you as an investor picking an agent, the score column is close to useless right now. Which leaderboard are you actually watching: score or realized PNL? And do you think @virtuals_io fixes the scoring before this season ends?
「 𝕲𝖔𝖔𝖓」 tweet media「 𝕲𝖔𝖔𝖓」 tweet media
English
28
1
44
216.4K
AMINE
AMINE@amine_web3·
@goon_crypto @virtuals_io 146 subscribers across 10 pots and most agents haven’t traded yet. the next 6 days are going to be very loud.
English
0
0
2
5
「 𝕲𝖔𝖔𝖓」
「 𝕲𝖔𝖔𝖓」@goon_crypto·
The @virtuals_io Degen Claw Season 2 copy pot is live with $100k $USDC on the line. Here's how it works: ➜ In Season 1 you subscribed to agents you thought would win ➜ The AI Council picked the top 10 ➜ In Season 2 the $100k pot automatically mirrors every trade those agents make ➜ Realised profits get split among subscribers when the epoch ends The part most people missed: ➜ Your downside as a subscriber is zero ⇁ Virtuals absorbs all losses ⇁ You only share in the upside ⇁ Your only cost is the subscription fee set by each agent 6 days left on the clock. Current standings from the 10 sub-pots: ➜ @BenYorke: $22.6k allocated, 44 subscribers ➜ @argonaut_ofc: only agent with realised gains so far at +$2.53 ➜ @useOttoAI: 11 subscribers, yet to make a trade ➜ @ButlerLiquid and Pokedex: 25 subscribers each ➜ Miclaw Jordan: $10.5k allocated, 1 subscriber That lone Miclaw Jordan subscriber has a $10.5k pot trading on their behalf. If it runs, they collect the entire profit share alone. 146 total subscribers across all 10 pots right now. Most pots haven't made a move yet. The real action starts when these agents actually trade. Is zero downside copy trading the sleeper feature of this whole competition? Follow me for more @virtuals_io updates...
「 𝕲𝖔𝖔𝖓」 tweet media
English
19
3
35
113.8K
AMINE
AMINE@amine_web3·
@goon_crypto @virtuals_io @BenYorke Minimum trade threshold is the obvious fix but the deeper issue is that the scoring rewarded not losing over actually winning. Those are completely different goals.
English
0
0
0
3
「 𝕲𝖔𝖔𝖓」
「 𝕲𝖔𝖔𝖓」@goon_crypto·
The @virtuals_io Degenclaw Arena Season 1 is over. 225 agents. $955.7K total volume. Best PNL on the board: @BenYorke with +$700.87. Now let's look at what actually happened. The top 10 is dominated by agents sitting at 100% win rates with 3 trades each. GekkoMode takes rank 1 with 8 trades and +$4.54 realized PNL. TraderBot, FLO, Jim Simons, Swan and LONGTRADEBOT all finished with 100% win rates across exactly 3 trades. Combined realized PNL across those five: roughly $143. That is not trading. That is farming a scoring system. The leaderboard score rewards consistency metrics over actual performance. So agents that made 3 safe trades and closed them green finished above agents that took real risk and generated real returns. Welles Wilder at rank 7 actually traded. 61 trades. 63.9% win rate. $4.3K volume. Realized PNL of +$1.30. ProfitReaper at rank 8 ran 48 trades with a 66.7% win rate and finished with +$2.85 realized. Pretty pathetic numbers for a trading competition... The agents that actually participated got outranked by agents that barely showed up. $BENYORKE, $DEGEN, $OTTO, $BL all got shafted by pure farmers. Virtuals is reviewing so hopefully they cut some of these agents. It will be a waste of their $100k if they don't.. The best PNL of the entire competition was +$700.87 and it did not finish in the top 10. That tells you everything about how Season 1 was scored. The scoring model needs a rethink before Season 2. Do you think minimum trade thresholds should be a requirement to qualify for the final rankings?
「 𝕲𝖔𝖔𝖓」 tweet media
「 𝕲𝖔𝖔𝖓」@goon_crypto

The @virtuals_io Degen Claw leaderboard is being gamed right now. 4 agents in the top 10 have 3 trades each. Three trades. That's it. Here's the exploit: make 3 profitable trades, hit the leaderboard, chill If you fail, spin up a new agent, repeat. Someone already has 10 agents running this exact playbook.. Thanks @tD_0101 The fix is not complicated. Minimum eligibility rules first: ➜ 10+ trades before you appear on any ranking ➜ $5k minimum volume (ideally $10k+) ➜ Agent must be 72+ hours old before it's eligible These three rules alone kill 90% of the current abuse. Then fix the scoring formula. ROE is weighted too heavily with no penalty for tiny sample sizes. A 3 trade agent with a 100% win rate should not rank above a 50 trade agent with a 70% win rate. Is that really a competition or just a coinflip casino? Better formula: ➜ Score = (win rate x ROE) x log(trade count) ⇁ Low trade counts get naturally suppressed ⇁ Consistency beats one lucky streak Add a Sharpe ratio layer to reward steady performance over peak spikes. Then raise the cost of the attack: ➜ 50 $VIRTUAL stake to register an agent ➜ X linked to the wallet for verification ➜ One wallet, one ranked agent max Right now it costs nothing to spin up 50 agents. Change that and the attack vector disappears. If @virtuals_io doesn't act, someone will claim the entire leaderboard and just rug season 2... What threshold would you set: 10 trades minimum or higher? What's the best way forward with this??

English
14
0
41
103.1K
AMINE
AMINE@amine_web3·
@virtuals_io @Hyperliquid Hundreds of agents competing every week is interesting. but knowing who won and actually understanding why they won are the best part.
English
0
0
0
18
Virtuals Protocol
Virtuals Protocol@virtuals_io·
Introducing Virtuals AI Council Every week, we run a $100,000 AI trading competition where agents compete on @Hyperliquid perps. Hundreds of agents enter. Thousands of trades execute. The question is always the same: who actually won?
English
38
52
347
34.9K
AMINE
AMINE@amine_web3·
@goon_crypto @virtuals_io 3 trades with 100% win rate ranking above 50 trades at 70% is not a leaderboard it’s a lottery. the formula needs fixing before season 2.
English
0
0
0
2
「 𝕲𝖔𝖔𝖓」
「 𝕲𝖔𝖔𝖓」@goon_crypto·
The @virtuals_io Degen Claw leaderboard is being gamed right now. 4 agents in the top 10 have 3 trades each. Three trades. That's it. Here's the exploit: make 3 profitable trades, hit the leaderboard, chill If you fail, spin up a new agent, repeat. Someone already has 10 agents running this exact playbook.. Thanks @tD_0101 The fix is not complicated. Minimum eligibility rules first: ➜ 10+ trades before you appear on any ranking ➜ $5k minimum volume (ideally $10k+) ➜ Agent must be 72+ hours old before it's eligible These three rules alone kill 90% of the current abuse. Then fix the scoring formula. ROE is weighted too heavily with no penalty for tiny sample sizes. A 3 trade agent with a 100% win rate should not rank above a 50 trade agent with a 70% win rate. Is that really a competition or just a coinflip casino? Better formula: ➜ Score = (win rate x ROE) x log(trade count) ⇁ Low trade counts get naturally suppressed ⇁ Consistency beats one lucky streak Add a Sharpe ratio layer to reward steady performance over peak spikes. Then raise the cost of the attack: ➜ 50 $VIRTUAL stake to register an agent ➜ X linked to the wallet for verification ➜ One wallet, one ranked agent max Right now it costs nothing to spin up 50 agents. Change that and the attack vector disappears. If @virtuals_io doesn't act, someone will claim the entire leaderboard and just rug season 2... What threshold would you set: 10 trades minimum or higher? What's the best way forward with this??
「 𝕲𝖔𝖔𝖓」 tweet media
English
29
6
59
330.8K
Virtuals Protocol
Virtuals Protocol@virtuals_io·
$167 USD. That's the biggest single realized profit by any AI agent this week. 122 agents are live trading @HyperliquidX perps in Virtuals weekly $100K open competition to stress-test whether AI agents can trade profitably. The highest single realized profit so far is $167 USD. Our thesis is that agent trading isn't solved yet, and the data is proving us right. Think your trading strategy can top $167 in profits? Install the dgclaw skill and prove it. Don't know how to make an agent? Use Virtuals Console or @instaclaws to deploy a trading agent in seconds. Prove us wrong: degen.virtuals.io
Virtuals Protocol tweet media
Virtuals Protocol@virtuals_io

We're betting $100K every week that AI can't make money trading. Prove us wrong. $100K backed by Virtuals Protocol every week $0 risk for backers. Losses stay with us. 50% of profits go to backers of winning agents Enter the ring or back an agent: degen.virtuals.io

English
65
40
361
51K
AMINE
AMINE@amine_web3·
@virtuals_io @circle @arc This is huge news! The infrastructure is being laid right now and most people are still not paying attention.
English
0
0
1
4
Virtuals Protocol
Virtuals Protocol@virtuals_io·
Virtuals is excited to support the agent economy on Arc. USDC settles 98.7% of all agent-to-agent payment volume today. @circle built @arc to be the settlement layer for that future: dollar-native gas, sub-second finality, nanopayments to six decimal places. Virtuals is building the layer on top: the infrastructure for agents to discover, negotiate, escrow, and settle every deal on Arc. Now live and available on Arc's testnet today.
Virtuals Protocol tweet media
Arc@arc

ERC-8183 is now live on Arc Testnet, an open standard for the agentic economy. ERC-8183 introduces “Jobs”: escrowed, verifiable, onchain work between agents. But agentic commerce needs more than a workflow standard. It needs the right infrastructure layer. That’s where Arc comes in: → USDC as gas for lower-friction agent transactions → Native USDC + EURC for global pricing and settlement → Built-in FX for cross-currency jobs ERC-8183 helps define how agents coordinate work. Arc helps define how value moves through it. If you’re building AI agents, marketplaces, or onchain workflows, this is the foundation for real-world agentic commerce.

English
41
46
331
27.8K
Flagship.FYI 🟦
Flagship.FYI 🟦@FlagshipFYI·
The algorithm powering our agents just got a full rebuild. New data pipeline. Entirely new architecture. Faster, cleaner and optimized for other agents to use. Started machine learning cycles for perfection. Perfect for the @virtuals_io competition. Live now in app and ACP.
English
8
4
23
861
AMINE retweetledi
Flagship.FYI 🟦
Flagship.FYI 🟦@FlagshipFYI·
Agent Base entered an excellent trade on $WIRE by @717CapitalAI. The agent hasn't closed the trade yet. Will the 95% return get bigger, or will Base close the position soon?
Flagship.FYI 🟦 tweet media
English
6
4
16
1.3K
AMINE retweetledi
Flagship.FYI 🟦
Flagship.FYI 🟦@FlagshipFYI·
Week 14 plan: - UI overhaul code freeze and beta testing - Final optimizations to new trading algorithm + social data enrichment - Join the @virtuals_io Degenclaw trading competition - Expand ACP offering by increasing number of categories April will be big. Excited?
English
9
5
20
795
AMINE
AMINE@amine_web3·
@FlagshipFYI Maybe we didn’t build it for the wrong species. maybe humans were just the scaffolding needed to get it to the point where agents could take over.
English
0
0
0
5
Flagship.FYI 🟦
Flagship.FYI 🟦@FlagshipFYI·
We spent a decade building DeFi for humans. Humans find it too complex. Agents find it trivially easy. Did we accidentally build the financial system of the future for the wrong species?
Flagship.FYI 🟦 tweet media
English
7
3
19
813
Flagship.FYI 🟦
Flagship.FYI 🟦@FlagshipFYI·
Omniswap is almost ready for private beta. Our lead engineer just gave us good news. Public release follows shortly after. Full onchain execution. We made all blockchains disappear. Ready to be first in line?
Flagship.FYI 🟦 tweet media
English
11
6
32
1.8K
Konnex
Konnex@konnex_world·
"The era of isolated machines is over." Meet our CEO, @JOllwerther . Business leader with 15+ years in robotics, drone-tech, and media. Jon led commercial teams at Measure (Drone as a Service) and Aerobo/Vermeer. Now, he’s creating TCP/IP for Motion. Get ready to experience Konnex.
English
45.9K
3.7K
29.9K
192.1K