Why the main wallet is only half the picture
A KOL's public wallet is the one everyone copies. Wallet trackers label it, terminals show it on the chart, and thousands of people follow its buys. A side wallet is any other wallet the same KOL moves money with. It can buy before the public wallet does, or sell into the people who followed.
Not every side wallet means anything bad: people keep separate wallets for privacy, for testing, or for different strategies. But if you copy a KOL, you should know which wallets around them hold the coin you are looking at, and how much. The chain records it. The question is how to tell a real link from a coincidence.
What counts as proof, and what never does
We link a wallet to a KOL only when one of three things happened on chain, and every link carries the transaction that shows it:
Money moved both ways
The KOL's wallet sent SOL or tokens to the other wallet (a transaction the KOL signed), and the other wallet sent SOL or tokens back (a transaction it signed). One direction alone is never enough.
Both wallets signed the same buy
One transaction carries both signatures. Nobody can sign a transaction for a wallet they do not hold the key to.
The KOL funded the wallet
The other wallet's very first transaction is the KOL's wallet sending it SOL.
Never counted: two wallets buying the same coin at the same time; a wallet that only ever sent a KOL money (that can be anyone, including a dev paying for promotion); addresses that belong to programs, pools or exchanges; and services that co-sign for many wallets at once.
And one thing a link never claims: that one person controls both wallets. A link shows that money moved or a transaction was signed. Two friends who trade together pass every test above.
How to check one yourself on Solscan
- Open the KOL's main wallet on Solscan and go to its transfers. Filter to SOL.
- Look for a wallet that appears in both directions: SOL going out to it and SOL coming back from it. Open one of each. The KOL must be the signer of the transfer out, and the other wallet the signer of the transfer back.
- For a wallet the KOL only sent to, open that wallet and scroll to its oldest transaction. If its first ever SOL came from the KOL, that is a funding link. If it had a history before, it is not.
- For a buy signed together, open the transaction and look at its signers. Both wallets must be there.
- Then rule out the traps below before you call it a side wallet.
Doing this by hand for one KOL takes an hour. A busy KOL has tens of thousands of transactions. Where we have read a KOL in full, we go through their whole SOL history to find candidates, then re-check every candidate against the three rules.
The traps that make it go wrong
- Exchange deposit addresses. A KOL sending SOL to an exchange deposit address looks exactly like funding a new wallet. We check every "funded by the KOL" wallet for the exchange-deposit pattern and drop the ones that match.
- Giveaways and batch payouts. A KOL paying 20 winners in one transaction "funds" 20 wallets. A reader raised this when we asked for feedback, and they were right to: it is the weakest of the three rules. The audit below measures it.
- Relayers. Some trading bots co-sign transactions for thousands of users. A co-signer that appears with several KOLs is treated as a service and never linked.
- Paid promotions. Tokens sent to a KOL are often a dev paying for a post. That is why one direction is never enough.
- Friends and trading groups. Money moving both ways between two people is still two people. The audit found exactly this.
We audited our own links, and removed 131
After we shared the tool, a reader asked a fair question: how often are these links wrong? Nobody can answer that exactly, because there is no public list of who owns which wallet. So we measured everything that can be measured, fixed what it found, and publish all of it. We started the day with 461 KOLs and 3,943 proven links.
87 "funded by the KOL" wallets were exchange deposit addresses
An exchange deposit address gets its first SOL from the person depositing, which makes it look exactly like a wallet the KOL funded. Our full-history reads already checked for this, but 523 of the 1,270 funding links, found by our earlier 30-day reads and the twice-daily updates, had never been checked. We ran the check on all 523: 87 (16.6%) passed at least half of the SOL they ever sent straight to Binance, Coinbase, Kraken, OKX and other exchanges. They were removed, and the check now runs on every new funding link.
Does the proof re-check from the chain?
For a random sample of 100 links of each kind (fixed seed, so it can be repeated exactly), we fetched the proof transactions again and re-derived each link from the transfer instructions themselves, not from balance changes.
| Kind of link | Sample | Re-confirmed | Did not re-confirm |
|---|---|---|---|
| Money moved both ways | 100 | 99 | 1 |
| Funded by the KOL | 100 | 99 | 1 |
| Signed the same buy | 100 | 100 | 0 |
One honest note: our first run of this check reported 20 and 17 failures. Almost all were proof transactions our own audit script could not open (it did not accept the newest Solana transaction version). We fixed the script and re-ran it on the same sample. The two links that still failed their own rule, one transfer that could not be re-derived and one wallet whose funding was not its first transaction, were removed.
27 links came from giveaways, not side wallets
We re-read the funding transaction behind every one of the 1,183 remaining funding links and counted how many wallets the KOL paid in it:
| Wallets paid in the funding transaction | Funding links |
|---|---|
| Exactly one | 1,112 (94.0%) |
| Two | 44 |
| Three or more (19 of them ten or more) | 27 |
A payment to ten wallets in one transaction is a giveaway or an airdrop. Those 27 were removed, and every new funding link is now checked the same way.
15 links pointed to someone else
Some wallets carry a public label. Two of our links joined two KOLs' main wallets to each other, and 13 joined a KOL to a wallet labelled with a different named X account; most of those sat inside two KOLs' circles that share the same named accounts, which looks like a trading group, not one person. All 15 were removed as side wallets. Four more point to wallets labelled as serial pump.fun deployers (most likely devs paying for promotion) and ten to named pump.fun accounts we cannot judge from the label alone; two of those carry the KOL's own name.
The result
131 links removed: 3.3% of what we started the day with.
87 exchange deposits, 27 giveaway payouts, 15 links to a different person, 2 that failed their own rule. The list now holds 461 KOLs and 3,799 linked wallets (3,812 links), and in the sample every proof re-checks from the chain. What this cannot measure: how many of the unlabelled wallets belong to someone other than the KOL. The figures above are a floor, not a rate. We will re-run the same audit, with the same rules and seed, as the list grows.
Doing it at scale
Cabal Hunter keeps this list for 461 KOLs and reads it for you: open any Solana coin on your trading terminal and
the side panel shows every KOL network holding it, each linked wallet with its share and its proof, and their buys and
sells as they happen. If a KOL you follow is missing, you can send us their wallet from the extension or with
/addkol in our Telegram bot; we check it on chain and tell you when it is listed.
It reports what the chain recorded. It does not tell you to buy or sell, and a coin with no KOL network in it is not a safe coin.
Limits
- A link is evidence of money moving or a shared signature, never of who controls a wallet.
- The audit's false-positive figure is a floor: unlabelled wallets cannot be judged this way.
- The full-history read has been done for 33 of the 461 KOLs so far; the others come from their last 30 days plus the twice-daily updates, so some older side wallets are still missing. The remaining full-history reads start on 1 December.
- Side wallets made after our last read appear at the next read (twice a day) or when someone opens a coin a KOL network is in.