Example case. Addresses and numbers are illustrative.
Eight eyeson every token.
Ocellus traces a token’s holders, follows their funding and checks the team’s claims against the chain. Eight agents each watch one thing, and a Skeptic challenges every finding before it reaches your report.
Robinhood Chain and Solana.
Scroll to watch an investigation
It remembersevery case.
Every scan is kept as a case: what was found, what was flagged, and what turned out to matter. When a case later shows something the scan missed, Ocellus knows exactly where it fell short.
Misses becomenew checks.
The missed pattern is turned into a check and replayed on past cases, to see what it would have caught and what it would wrongly flag. Then every scan runs it, so the next token gets a sharper look than the last.
Eight eyes.One argument.
Paste a token from Robinhood Chain or Solana. Eight agents each watch one thing: holders, funding, trades, the website, X and the promises made. They share what they see and challenge each other until only what the evidence supports is left.
Robinhood Chain0x4a1d9c07e2b5f38a61c0d7e93f
Holder Analyst
Who holdsthe supply?
Every transfer since launch is replayed and every balance checked against the chain. The ten biggest wallets hold just 3.1%, so ownership looks spread out.
holders found
1,108
held by the top 10
3.1%
Funding Investigator
Are these 54 walletsconnected?
The Funding Investigator disagrees with that picture. Fifty-four smaller holders bought through the same router within forty minutes, each from a wallet with 15 to 25 earlier transactions. Together they hold 9.8%.
wallets bought the same way
54
held between them
9.8%
History Analyst
Is the volumereal?
Every buy and sell is tied to its wallet. Most wallets loop out of the pool and straight back in, so the busy chart is mostly the same money going round.
of traders sold back out
71%
of volume came from them
84%
Web Investigator · Social Analyst
What does theproject say?
The website, its domain record and the X account are lined up against the launch. The domain was registered three hours after the token, which went straight onto Uniswap with no launchpad.
hours from launch to domain
3
launchpads used
0
Promise Auditor
Does the chainagree?
Each promise is checked against the records. Liquidity locked: confirmed, the pool tokens are burned. Owner renounced: confirmed. A community-owned supply: doubtful, given those fifty-four wallets.
promises checked
6
doubtful
1
Skeptic · Funding Investigator
Then theyargue.
The Funding Investigator calls the fifty-four one group. The Skeptic points out that thousands of people use that router. The evidence settles it: a strong pattern, reported as a pattern, never as proof of one owner.
challenges raised
12
findings downgraded
3
Lead Investigator
What survivesis your report.
It opens with the red flags and what looked fine, keeps what survived the argument, and links every sentence to its record.
Lead Investigator’s summaryIllustration
Several red flags
Ownership looks spread out, but most trading is wallets going in and out, and 54 wallets bought the same way. The pool’s liquidity is burned.
Record
Every token transfer
Wallets
1,944 of 2,738
Still holding
38
Record
First buys
Window
40 minutes
Earlier transactions
15 to 25 each
Record
Pool created
Block
58,406,131
Transaction
0x51ab…c9e0
Record
Contract read
Block
58,412,902
Agent
Promise Auditor
Sharper withevery scan.
Any token on Robinhood Chain or Solana. Eight eyes on every finding, and every case makes the next look sharper.
Each check exists because a reviewed case showed the scan missing something. Reports are kept, reviewed cases turn into checks, and a check is tried on earlier cases before it ships. Recognising a wallet across separate cases is still to come.
What prompted itOwnership looked spread out, but dozens of small wallets had bought the same way.
The checkMatching buys: same route, same burst of time, similar wallet history.
Runs in every scan
What prompted itTrading looked busy, but most wallets sold straight back out.
The checkBuy-and-sell churn, measured across every transfer since launch.
Runs in every scan
What prompted itThe website promised locked liquidity and a renounced owner, and the scan could not tell.
The checkPool liquidity burn and the contract owner, read at the snapshot.
Runs in every scan
What prompted itA wallet from a flagged case can turn up in the next token, unnoticed.
Planned checkMemory across scans: wallets from earlier reviewed cases are named when they appear again.
Not built yet
Why it is called Ocellus.
An ocellus is a simple eye. A spider has eight of them, each watching a different direction, and no single one sees the whole scene.
It is also the eye on a peacock’s feather. There is no blue or bronze pigment in it: the colour comes from structure too fine to see. Ocellus reads a token the same way. Eight agents each watch one part of it, and the report describes the structure underneath, not the surface.
What it will never tell you.
Who owns a wallet
Shared funding, timing or buying patterns can suggest a link. Ocellus never presents it as proof that one person controls two wallets.
Whether a token is a scam
It counts red flags and says what looked fine. It never labels a token, a team or a wallet a scam.
That no red flags means safe
A clean report means nothing stood out in what public records show, not that a token is safe to buy.
That a learned check is always right
Every new check is replayed on past cases before it runs, and its findings still carry their strength and caveats.
Where the price goes next
No predictions and no trading signals. Ocellus never requests token approvals or transfers; wallet sign-in uses a message signature.
That a gap does not matter
When a data provider or scope limit stops a step, the report says what was not covered.