Ex-dev turned onchain detective π΅οΈ Exposing scams. Tracking wallets. Calling memes.
Onchain
Joined May 2024
- Tweets997
- Following136
- Followers1.5K
- Likes716
Pinned Tweet
My AI trading bot made +9% on day 1.
Ten days later it was down 54%.
2,700+ buys. 18 runs. Every trade logged.
Here's what nobody posting those terminal videos will show you π
Update: I never turned the bot off. π€
The article covered 4 days. It's now been 11.
7 strategies. $70,000 of paper money.
3,835 buys.
Left: $36,900.
Still not one green day.
My AI trading bot made +9% on day 1.
Ten days later it was down 54%.
2,700+ buys. 18 runs. Every trade logged.
Here's what nobody posting those terminal videos will show you π
1
57
The craziest number from 11 days:
716 trades crashed β75% or worse.
They lost $34,686.
The whole account lost $33,093.
Every other trade combined was actually green. A few crashes ate everything.
Median time to the crash: 6.5 minutes.
1
3
I wrote down how the first 4 days went, step by step:
the +9% day 1, the backtest that lied, the 56% win rate that lost money.
Full article π
teapawt.net/MisterFeeman/status/21β¦
NFA. Paper trading. Data from GMGN π gmgn.ai/r/bytwk77l
My AI trading bot made +9% on day 1.
Ten days later it was down 54%.
2,700+ buys. 18 runs. Every trade logged.
Here's what nobody posting those terminal videos will show you π
2
The tell was there the whole time.
At $15β20M market cap, $COGE was trading $50β200K an hour.
A "$20M coin" that can't absorb $100K of selling isn't a $20M coin.
Before you buy, compare market cap to liquidity and volume. Takes 10 seconds.
$COGE
CA: 8UQdy5sHMVWUWSHeA5B3ZuuAVNumjSLPXdf19MXcpump
1
18
My bot bought tokens like this for 4 days.
280 exits at β75% or worse. Median time: 5.8 minutes.
A stop-loss can't help when the price drops 70% inside one candle.
Full data π
teapawt.net/MisterFeeman/status/21β¦
NFA π gmgn.ai/r/bytwk77l
My AI trading bot made +9% on day 1.
Ten days later it was down 54%.
2,700+ buys. 18 runs. Every trade logged.
Here's what nobody posting those terminal videos will show you π
9
solana:75VDY6P9kJyAp7QGYRzq7o85oKBSXpR9XmaWVHcspump was 90 minutes old. $293K market cap. 2,051 holders. π
Then one candle.
$293K β $13.5K in 60 seconds.
Now: $4.4K. Down 99% from its $699K high.
Nobody had time to sell.
1
99
What the dev wallet did:
17:46:40 β sold $695 worth
17:47:18 β pulled the liquidity
17:47:21 β sold 418M tokens for $2,018
The wallet now holds 0 SOL.
solana:75VDY6P9kJyAp7QGYRzq7o85oKBSXpR9XmaWVHcspump
CA: 75VDY6P9kJyAp7QGYRzq7o85oKBSXpR9XmaWVHcspump
2
29
My bot tracked exits like this for 4 days.
280 exits at β75% or worse.
Median time after the buy: 5.8 minutes.
A stop-loss doesn't save you when the price drops 95% inside one candle.
Full data π
[ΡΡΡΠ»ΠΊΠ° Π½Π° ΡΡΠ°ΡΡΡ]
NFA π gmgn.ai/r/bytwk77l
10
My backtest said +3.4%.
The same strategy live: β33%.
The backtest only saw tokens that survived. Every dead token was missing from the data.
Most bot screenshots on your timeline have the same hole π
My AI trading bot made +9% on day 1.
Ten days later it was down 54%.
2,700+ buys. 18 runs. Every trade logged.
Here's what nobody posting those terminal videos will show you π
1
1
71
Yesterday: the $TAPE dev turned $8.2K into $60.3K in 116 seconds.
Today:
$189K β $2.1K
726 holders β 97
Bots bought it for you. Then they left.
I built one of those bots. It lost 54% in 4 days.
Here's everything it taught me π
My AI trading bot made +9% on day 1.
Ten days later it was down 54%.
2,700+ buys. 18 runs. Every trade logged.
Here's what nobody posting those terminal videos will show you π
2
94
The buy went through the whole remaining bonding curve in one transaction: 793M tokens, $8,206.
The first sell came 11 seconds later.
76 sells in under 2 minutes. $60,340 out.
Who was buying?
GMGN flags 89% of $TAPE traders as bots.
99 of the top 100 holders are wallets, and they all bought within 4 minutes. 78 are tagged bundlers.
1
25
Bots watch for fresh migrations and buy without asking why a 7-day-old dead token woke up.
I know, because I built one.
My bot's result: β54% in 4 days. 280 exits at β75% or worse.
The full data is in the article π
[ΡΡΡΠ»ΠΊΠ° Π½Π° ΡΠ²ΠΎΡ ΡΡΠ°ΡΡΡ]
$TAPE CA: 7xP2r4Vtjy5opBZjbPcCc4N7BkRhhQddqsFQDpDopump
NFA π gmgn.ai/r/bytwk77l
19