u8 Profile
u8

@atarashi

Followers
789
Following
2K
Media
37
Statuses
237

co-founder @0xGhostLogs

Joined February 2009
Don't wanna be here? Send us removal request.
@atarashi
u8
11 days
link to dashboard:
Tweet card summary image
sandwiched.me
Experiments
0
2
28
@atarashi
u8
11 days
You might have heard about "Prop AMMs" on Solana. We built a lightweight dashboard:. - streams realtime data from 2 Prop AMMs.- plots continuous-depth curves => can see sub 1 bps.- cross-pool router simulator => can see sub 0.5 bps
26
29
247
@atarashi
u8
14 days
New weekly Claude Code limits dropped. They sent out 3 variations of the email depending on which plan you have so summarized them here. Based on this, Max 20x isn't really 4x the Max 5x plan
Tweet media one
5
3
19
@atarashi
u8
17 days
One interesting thing about GoonFi that I haven't seen anyone mention (or discover?). GoonFi uses both whitelist and blacklist controls, something most Solana programs don't. * Whitelist: Only accepts calls from Jupiter v6.* Blacklist: seeing 246 traders already flagged. An.
@SlippageMonster
Slippage Monster
19 days
1000 *genuine* likes and I tell the world who GoonFi and HumidiFi are! . Gobble Gobble.
2
0
14
@atarashi
u8
20 days
Indeed! And you can confirm for yourself (*). $ ./target/release/solfi-sim spreads 1 . (*) Using the solfi-sim repo
Tweet media one
@HypoNyms
Ben ⌛
20 days
Daily reminder that spreads on Solana are sub-1bp! . Use DeFi.
4
1
22
@atarashi
u8
20 days
RT @chrischang43: Our new Solana mainnet validator Ghost is live. If you find our public data & research useful, please consider delegat….
0
10
0
@atarashi
u8
25 days
If you don't believe in wide(*) sandwiching, here's a thought experiment:. Let's assume a MEV bot:. (1) Has access to a transaction, with or without the signature. (2) Does _not_ have privileged access to the current leader _and_ cannot bundle through Jito. So you're either
Tweet media one
@bertcmiller
@bertcmiller ⚡️🤖
3 months
I am highly skeptical of "blind sandwiching". It seems far more plausible that "blind" sandwiching is simply validators sandwiching and trying to avoid being caught.
1
1
5
@atarashi
u8
26 days
This might be controversial, but your transactions should be able to fight back against malicious sandwiching validators. I built a simple program to do exactly that. You can't know at runtime if slippage is natural market movement or a sandwich attack. But if your swap lands
Tweet media one
9
11
63
@atarashi
u8
3 months
RT @chrischang43: 2/ 16 months of blocks, 8.5 Billion trades, over $1T in DEX volume. Over this time period, the Atomic Arb volume was $3….
0
1
0
@atarashi
u8
3 months
Skimmed 10 "comprehensive" and "in-depth" guides about eip-7702. Not a single one had an example on how to actually build a 7702 transaction. The only one I found was in the alloy examples repo. Take a look:
Tweet card summary image
github.com
Example code for using alloy and alloy-core. Contribute to alloy-rs/examples development by creating an account on GitHub.
1
0
5
@atarashi
u8
3 months
Good time as any for my yearly reminder that 550,000 ETH is still locked away forever in original Parity multisig wallets. If there's a vulnerability in Gnosis Safe or other smart contract-based wallets, it would be foolish to expect a chain fork.
@gakonst
Georgios Konstantopoulos
3 months
What makes people comfortable storing large $ in Gnosis Safe? . What would it take to replicate that in another smart contract that acts as a user wallet?.
0
0
5
@atarashi
u8
4 months
Cool use of litesvm to simulate SolFi swaps.
@chrischang43
chris
4 months
Nearly 25% of atomic arbs (and 13% of rev) on Solana touch a SolFi pool. I was curious how SolFi works. When something is closed source and with no IDL sometimes the best thing to do is to simulate. Over the weekend, my co-founder wrote a simple simulator that fetches all pool
Tweet media one
0
0
7
@atarashi
u8
6 months
RT @chrischang43: I created a DATASET of all $libra trades so you can do your own analysis. 6 differente AMMs.67 distinct pools.117k unique….
0
4
0
@atarashi
u8
6 months
Even better, use a Dutch auction. Bonding curves still rely on a party (not the market) trying to 'solve' price discovery at the wrong level—tweaking constants or formulas isn’t the answer. Let the market decide the price, not a fixed curve. *then* switch to an AMM.
@AustinAdams10
Austin Adams
6 months
creating a steeper bonding curve is a way to solve this problem. It should be EV- to snipe every launch or it will continue to happen. Doppler-v3 utilizes a programmable bonding curve that controls the steepness of the curve by adjusting 1 parameter.
1
1
5
@atarashi
u8
7 months
onchain thinking machines (whether they use a TEE or not is an implementation detail) are the only way under-collateralized loans will ever work in crypto. an AI agent with a verifiable brain can be perfectly risk-scored, creating trustless, scalable credit systems.
1
0
3
@atarashi
u8
7 months
RT @Teknium1: Unbelievable the amount of cope, seethe, and hoop jumping people are doing to discredit deepseeks accomplishments lol.
0
222
0
@atarashi
u8
7 months
RT @HypoNyms: The fundamental flaw with swQoS is that the top validators are NOT the top tx senders. It doesn't even maximize token value….
0
12
0
@atarashi
u8
7 months
Once again, impressive work by the fuzzland (now: solayer) team. Within 5min of the hacker's implementation contract being deployed, they were able to detect the vulnerability and bridge over funds to initiate a rescue. Building and submitting the rescue txn itself seems to.
@tonykebot
Tony KΞ
7 months
We just automatically hacked the hacker and rescued 1.4M USDC!. 100% of fund were returned to the project owner. > 🧵 Here's how the hacker is whitehat-hacked
Tweet media one
Tweet media two
0
1
3
@atarashi
u8
7 months
Remember Chi gas tokens?. The original gas arbitrage. You could mint or buy them when gas was cheap and then use them when gas was expensive. Burning Chi gas tokens in a transaction would refund gas, up to 50% of the total gas usage.
1
0
6