A RobinX token report (robinx.io/t/{address}, or GET api.robinx.io/report/{address} for a bot) reads one Robinhood Chain token from public chain data and public X posts. It is a set of measurements with the rules that produced them, not a rating of the project. This page says what each part means, part by part.
One word with its reasons, computed from the deployer's other launches and two present-state rules: the token under judgment is left out of the record it is judged on, so a token never vouches for itself.
| Verdict | Meaning |
|---|---|
trusted | The deployer's record scores 70 or more, which needs at least two other launches that reached real activity. One launch that traded heavily is not a track record. |
mixed | Neither trusted nor avoid: a score from 40 to 69, a score held at 69 because fewer than two other launches reached real activity, or a reading moved down by a present-state rule. |
avoid | The token, or another launch by its deployer, came through a known launch template (in every sampled launch through it, 342 of 342, the deployer removed all of the liquidity, a median of about eight minutes after adding it), or the deployer's record scores under 40. |
serial_spammer | Ten or more launches by the deployer, this one included, and none of the others reached real activity. |
new_deployer | No other graded launch. Nothing is known about the deployer yet, which is a fact about the record, not about the token. |
deployer_unknown | No deployer record applies: no deployer is attributed for the launch path yet, the token arrived through a cross-chain relayer (a bridge, not a person), or the record could not be read on this request (the answer then carries degraded: true). |
tokenized_stock | A genuine Robinhood stock or ETF token: a market, not a launch. Price, volume and calls are measured; no deployer or custody judgment applies. |
Two present-state rules can move a verdict down to mixed: one wallet outside the pools holding 25% or more of the supply, or the ten largest holding 60% or more (applied to trusted and new_deployer), and a deployer that distributes supply wallet-to-wallet (applied to trusted). A look-alike named like a stock token is judged as the launch it is and flagged named_like_robinhood_stock_token.
Every launch by the same wallet: when, how many swaps, how much WETH volume, how many traders, and whether it reached real activity (at least 1 WETH of cumulative volume and 20 distinct traders). The score, 0 to 100, is computed from that history by transparent rules, and every score returns machine-readable reasons. A deployer's single success is capped at 69; a deployer with any launch through a known launch template is capped at 20; a cross-chain relayer is never a deployer. The out-of-sample test, a deployer judged only by its earlier tokens against its next token's outcome, is published live at robinx.io/proof. The measure can be evaded by rotating wallets: a fresh wallet has no record and reads new_deployer, a gap, not a pass.
Price and market cap come from the current state of the pool holding the most WETH, or WETH-equivalent for a USDG-quoted pool, in range at the current price, across Uniswap v2, v3 and v4; when the two deepest pools disagree on the price by more than 10%, the deepest one that traded in the past day is used. A USDG-quoted pool is converted at the USDG/WETH rate read at the same time. A token whose deepest pool holds under 0.05 WETH shows no market. A price read more than three days ago is not shown as current. The line says which quote priced it and when the pool was read. The WETH price uses no oracle; the dollar figure converts it at a public ETH/USD spot price. Depth in range is what a trade meets at the current price, not the capital behind the token.
On Alpha and Omega keys the report API (the field lp) and the one-call check carry who holds the live liquidity positions of the token's main pool, read from chain state: wallet (a plain address holds a live position), contract (every live position is held by a contract), burned, withdrawn (no live liquidity left) or unknown (nothing resolvable; a Uniswap v2 main pool reads unknown). It is an observation about custody, never a prediction. The public page does not show it. The method and its limits are on the LP custody page.
The first X accounts RobinX captured naming the token, in time order, each with its own measured record: graded calls, early rate, median one-hour move and flags, under the rules on the grades page. Whether a given call came before or after the move is on that account's receipts; the report does not count followers as evidence.
Off-market recipients: wallets that received the token by transfer rather than purchase, how much, how many blocks after launch, and whether they sold into the pool. On the genesis corpus this insider flow was where the chain's largest extractions came from, which is why it is a first-class input. The holder-structure table, on plan keys, labels the wallets that moved supply with RobinX's own insider and deployer records and, where the owner has published one, an ENS name or a Farcaster account; wallets verified by one Farcaster account are shown as one holder. RobinX adds no identity beyond what the owner published on chain.
A live forward test of the deployer measure: since 10 July 2026, every launch with an ETH-quoted Uniswap v3 pool whose deployer had two or more tokens reach real activity, by the deployer record read at entry, is entered with paper about 30 minutes after launch and held. If the token is in the basket, the report shows when it entered, the deployer's score then, and its current and peak multiples of the entry price. Paper results are a test of the measure, not a track record and not a recommendation.
The chain readings on the page are reproducible from the chain and the posts; the dollar figures convert them at a public spot price. For a bot: GET api.robinx.io/report/{address} ($0.05 over x402, or free with ?tier=free at 25 a day per IP, or on any plan key); GET api.robinx.io/check/{address} is the one-call pre-trade checklist (clear, blockers, warnings and the readings; 100 free a day per IP, $0.02 over x402). Every answer carries its rules version and timestamps.
What a report is not. Not a rating of the project, not a prediction, not advice. A verdict describes the deployer's record and the token's measured structure under published rules; a new or unknown deployer is a gap in the record, not a judgment. Labels such as avoid or serial_spammer describe an address's record under the rules; they are not accusations against any person. The strongest claim on the page is a measurement with its receipt.
trusted (a score of 70 or more with two or more other launches that reached real activity) nor avoid (a known launch template, or a score under 40), or a present-state rule moved the reading down. The reasons beside the verdict say which rules applied.