How RugWatch judges every pump.fun meme at birth
RugWatch watches every pump.fun launch (about 50,000 a day) and sorts it using facts measured on the Solana blockchain, with the same rules for everyone. A verdict describes an observed behaviour, not a person. It is neither financial advice nor an accusation.
What RugWatch does, and doesn't do
- ✅ It reads public data only: Solana transactions, pump.fun accounts, and the links the creator declared (website, X, Telegram).
- ✅ It applies the same rules to everyone, automatically, about 1 second after birth.
- ❌ It never says who is behind a wallet and never names anyone.
- ❌ It does not predict the price. A good verdict is not a reason to buy; a bad verdict is not proof of fraud.
🥚 The incubator: the verdict at birth
Every meme spends 2 minutes in the incubator. The data comes in step by step (creator after 1 s, first buys after 30 s, full analysis after 90 s) and the verdict can change along the way. The first rule that matches decides.
20 or more launches in 7 days by the same creator or the same clan.
The creator (or its clan) launched 5 or more tokens in 7 days and none of them graduated. The automatic flag also requires that at least 3 of those tokens died within the hour.
At least 3 creator wallets funded by the same wallet (the first SOL each one received). Exchanges are excluded.
The token shows another token's address, the website of a tech giant or a media outlet, a website already used by an earlier token, a website used by 10 or more tokens, an X account already used by another creator, someone else's tweet instead of its own account, or the official X account or website of a known brand. A borrowed profile is set aside for 7 days.
50 % or more of the first buys come from bots (see below).
After 45 seconds, less than 0.4 SOL has come in: nobody is buying.
The creator's first known launch, outside any clan, with at least one link that truly belongs to it: a website that talks about the token, a real X account (not just a tweet), or a Telegram group.
The creator's first known launch, outside any clan, but with no website and no X account that can be checked.
The creator has launched other memes before without being a factory. The card shows how many they launched and how many graduated: look at what they did before.
🤖 Bots, squads and packs
A wallet that buys then sells within 15 minutes, on 20 or more different tokens in 24 hours.
3 or more wallets buying the same token, in the same second, for the same amount.
Bots that buy AND sell together (within 3 seconds) on at least 8 tokens, with at least 30 % of their trades in common.
A token whose buys come largely from bots (40 % or more), or that is hit by at least 2 squads. The card always shows the measured share of bot buys.
📊 What the figures show
Real RugWatch measurements, night of 9 to 10 October 2026 (about 28,000 births). “Graduated” means the token filled its pump.fun bonding curve.
| Profile at birth | Memes | Graduated |
|---|---|---|
| 😇 First launch with its own links (✅) | 982 | 4.9 % |
| 🧑🌾 First launch, no proof (⚠️) | 7,325 | 4.6 % |
| 🎰 Meme machine (⛔) | 4,751 | 2.3 % |
| 👤 Known creator (⚠️) | 3,307 | 1.9 % |
| 🔒 Serial creator (⛔) | 10,310 | 0.5 % |
| Share of buys made by bots | Tokens | Graduated |
|---|---|---|
| less than 10 % | 724 | 8.7 % |
| 50 % or more | 725 | 1.8 % |
Takeaway: a first launch graduates about 10 times more often than a serial creator's meme, and a token bought mostly by bots almost 5 times less often than one bought by humans. The creator's record and the presence of bots say far more than the chart. Even the best groups rarely graduate: pump.fun remains a lottery, and a good creator guarantees no price.
⚠️ Limits
- An automatic rule can be wrong: a real beginner who reuses a website by mistake, a wallet funded by a friend who also funds other creators…
- Some data arrives late (domain age, the funder of a very active wallet). The card shows which checks are done: creator, first buys, full analysis.
- The rules change when the measurements show they are wrong. The figures above will be updated.
⚖️ Contest a verdict
You think a verdict is wrong (for example, your website was flagged as “already used” although it is yours)? Explain it here. Every request is reviewed by hand, and nothing you send is published.
🛡️ Guardians
The incubator, the verdicts, the creator cards and the village are free for everyone, with births shown 60 seconds late. Guardians, the holders of APO (the Aporius token, launching soon), only unlock speed: every birth live, an alert for every serious creator, and opening everything in one click. You prove it by signing a message with your wallet: no transaction, nothing is spent.
Aporius