For agents · API

The same scanner, as JSON.

Trading bots and AI agents can call the scanner directly. Plain HTTPS GET, JSON out. A free key is optional and raises your limit. Verdicts come from on-chain data and code, not from a language model, so the same token gives the same answer.

Endpointsall GET, JSON

GET /api/scan?chain=sol|base|rh&ca=<address>

Full risk scan of one token: score 0–100, verdict, caps, every check with its value, pools, top holders, bundles, deployer.

20 fresh scans per 10 min per IP; a repeat within 60 s is cached and free

POST /api/keys {"label":"my bot"}

Free API key, shown once. Send it as the x-api-key header on /api/scan: 500 fresh scans per day per key instead of the per-IP limit.

3 keys per IP per day

GET /api/search?q=<ticker>&chain=

Find tokens by $TICKER. Sorted by real liquidity, fake-quote pools flagged, likely original marked.

60 s CDN cache

GET /api/radar

Fresh launches our radar scanned in the last 24 h, delayed 10 min.

public

GET /api/track

Track record: every live verdict, its price 24 h later and whether it held.

public

GET /api/proof?hash=<64 hex>

Merkle path from a verdict hash to its 15-minute root. Without hash: the latest roots.

public

GET /api/status

When the background ticks ran and how long they took.

public

GET /api/aim

$AIM supply, burned amount and price.

public

Reading a scanshortened

{
  "chain": "SOL", "ca": "EKpQ…zcjm", "symbol": "WIF",
  "score": 95, "verdict": "CLEAR", "confidence": 90,
  "caps": [],
  "checks": [
    { "id": "mint", "group": "contract", "label": "Mint authority", "value": "revoked", "sev": "ok" },
    { "id": "lp", "group": "liquidity", "label": "LP lock", "value": "…", "sev": "ok" }
  ],
  "market": { "price": 1.9, "liq": 4200000, "fdv": 1900000000 },
  "ms": 2400, "at": 1760600000000
}

Rules for agents

verdict is CLEAR (75+), CAUTION (40–74), HIGH RISK (16–39), AVOID (0–15) or TOO EARLY (pair under 30 min, no red flags yet). LOW DATA (under 60% of checks ran, no red flag found: not a warning, try again later) and CORE ASSET (stablecoin or wrapped coin, no memecoin verdict) have no meaningful score.
confidence is the share of checks that could run. Under 60 the score is held below 60.
caps explain why a score was limited. Any cap at 10 or below means do not buy.
sev per check: ok, warn, bad, info. Info checks do not move the score.
On HTTP 429 wait a few minutes. Cache results on your side for at least 60 s.
Holders, linked-wallet clusters and the deployer are in every scan response. Paid Pro data (X mentions, deployer track record) is not in the free API yet.