Four checks. Facts only, no advice. Overview · OpenAPI · llms.txt
Machine-readable spec at /openapi.json (OpenAPI 3.1). Most agent
frameworks will turn that straight into a callable tool. Payment terms at
/.well-known/x402.
Is this yield pool what it claims to be?
| Param | Meaning |
|---|---|
pool | DefiLlama pool id, from yields.llama.fi/pools |
Prices every underlying asset against its peg, compares APY to its own history, reports whether the yield survives once emissions stop, and checks exit depth.
curl "https://poolproof.ru/free/pool?pool=<uuid>"Flags: DEPEG OPAQUE SPIKE UNSTABLE EMISSION THIN
What does leaving this Solana position actually cost, at your size?
| Param | Meaning |
|---|---|
mint | Solana token mint address |
size | Position size in whole tokens. The whole point is that the quote is for your size. |
Quotes your real size against a live router, then compares it to the price at negligible size. The difference is what leaving actually costs.
curl "https://poolproof.ru/free/exit-cost?mint=DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263&size=50000000"Flags: SIZE_WALL EXIT_COST
Is this chain alive, or has the capital left?
| Param | Meaning |
|---|---|
chain | Chain name as DefiLlama spells it |
curl "https://poolproof.ru/free/chain?chain=Bitlayer"Flags: COLLAPSED STILL_FALLING TINY
Who launched this pump.fun token, and what became of their previous ones?
| Param | Meaning |
|---|---|
mint | Solana mint address of the pump.fun token |
Returns how many tokens the same creator wallet has launched, how many of them reached the exchange, the graduation rate of that creator group against the overall base rate, and a flag when the creator filled the bonding curve with their own purchase — in which case reaching the exchange reflects their buy, not demand. Price, volume and liquidity are public everywhere; that a token is the 238th from a wallet whose tokens have never survived is not.
Built from our own recording of every pump.fun launch and its author since 10 August 2026. A wallet active before that date appears unseen, and the response says so rather than calling it new. Empty flags mean nothing failed these particular checks, not that the launch is safe.
curl "https://poolproof.ru/free/launch?mint=<mint>"Do we hold a research note on this that no data feed reports?
| Param | Meaning |
|---|---|
q | Asset symbol, protocol name, chain name or contract address |
44 notes covering dead redemption bridges, tickers shared by unrelated projects, points programmes that ended quietly, marketplaces that shut down, roles whose real capital gate appears in no documentation, and measurement failures that make a losing system look profitable.
curl "https://poolproof.ru/free/known?q=renBTC"Verdicts: unredeemable ticker_collision stale_data
token_issued programme_ended shut_down gated
no_edge leak bug method hft_only exhausted
An empty result means we have not investigated it. It is not a clean bill of health.
Swap /v1/ for /free/. No key, no signup, ten calls an hour per address.
Same answers, so you can decide whether it's worth paying before you pay.
Call a /v1/ path with no payment and you get 402 with the terms in the body:
{
"x402Version": 1,
"accepts": [{
"scheme": "exact",
"network": "base",
"maxAmountRequired": "5000",
"asset": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913",
"payTo": "0x11Efa031D56316bD63720d172f9d5c1448e9a21F"
}]
}Sign the authorisation, base64 it into an X-PAYMENT header, repeat the request. Any
x402 client library does this for you. Settlement is USDC on Base and
takes about a second.
Every response carries a flags array. Empty means nothing failed. Each flag has a
code you can branch on and a msg for humans. Nothing is scored or ranked,
because the decision isn't ours to make.
What we record — live counts behind /v1/launch.