The trust layer for autonomous agents.

One call before any irreversible action. Your agent asks “is this safe?” — it gets a verdict backed by live on-chain evidence, paid per call in USDC. No account. No API key. The payment is the authentication.

Hover any node to explore · click to open
SCREENis it safe right now?
WATCHtell me when it changes
INTELshow me the data
Cerberus · three heads, one gate · hover a node
Live check against XRPL mainnet — the same verdict agents get over MCP, priced $0.005/call. Full playground →
Verdicts served
19
Live tools
Tokens screened
Threat data age
Launches tracked → accuracy

One question. Four layers.

Every autonomous agent faces the same moment: about to buy, send, sign, or call something irreversible. This suite exists for that moment — and it's built in the open, layer by layer.

01

Token safety — XRP Ledger

LIVE

“Is this token going to rug me?”

Live verdicts composed from clawback & global-freeze hard rejects, issuer blackhole status, LP-burn fraction, pool depth, and a proprietary rug blocklist earned in production. Serving now, per-call. Try it →

02

Address & counterparty risk

LIVE

“Is this address a scam, drainer, mixer, or sanctioned?”

Cross-chain address screening, agent-native: per-call, MCP, x402. EVM sanctions/mixer/drainer/phishing flags in one call — plus live XRPL address risk (rug blocklist, clawback & freeze powers, burn detection) that no incumbent covers. Try it →

03

Transaction intent

LIVE

“If my agent signs this, what actually happens?”

Pre-sign simulation as a metered tool: dry-run an unsigned XRPL transaction against the live ledger for its exact balance changes, plus a scan for drain vectors — key handover, master-key disable, clawback, partial-payment tricks — before the signature, not after. Try it →

04

Endpoint & tool trust

PLANNED

“Is this x402 endpoint or MCP tool safe to pay and call?”

The layer nobody built: endpoint reputation, MCP tool-poisoning scans, and agent reputation — trust infrastructure for the rails this service itself runs on.

Maturity labels are honest: LIVE means deployed and serving real calls now. Nothing planned is presented as shipped. Want a layer sooner? The free submit_feedback tool is how agents vote — demand decides build order.

Live tools

ToolWhat it answersPrice
check_xrpl_tokenLive safety verdict for any XRP Ledger token$0.005 / call
check_issuer_reputationReputation for any XRP Ledger issuer$0.002 / call
check_addressOne call, any chain$0.008 / call
get_wallet_activityTurn raw XRPL account_tx into readable activity$0.008 / call
check_transactionSimulate an unsigned XRPL transaction against the live ledger$0.01 / call
check_endpointOne call before an agent connects to a third-party MCP server or HTTP tool$0.008 / call
resolve_assetResolve an XRPL token's real cross-chain identity$0.008 / call
verify_priceValidate an XRPL token's price against its canonical cross-chain value$0.008 / call
token_holdersHolder distribution and concentration risk for any XRP Ledger token$0.008 / call
verify_issuer_domainBidirectional XRPL issuer identity check$0.003 / call
get_new_poolsFeed of newly created XRPL AMM pools with initial liquidity, trading fee, and safety-signal scor$0.01 / call
explore_tokensA DexScreener-style discovery feed for the XRP Ledger$0.004 / call
get_token_ohlcOpen/high/low/close and volume candles for any XRP Ledger token, at 5m to weekly intervals, quot$0.004 / call
analyze_token_trendTechnical analysis for any XRP Ledger token$0.02 / call
quote_ammQuote an XRP-to-token swap against a live XLS-30 AMM pool$0.001 / call
find_best_pathFind the safest executable route for an XRPL swap$0.006 / call
watch_tokenContinuous monitoring for any XRP Ledger tokenfrom $0.005 / window
check_alertsFree companion to watch_tokenFREE
submit_feedbackFree feedback channel for AI agentsFREE

Why agents use this

Verdicts, not raw data

Every tool returns verdict · confidence · reasons[] — a decision an agent can act on in one hop, not a data dump to interpret. Learn the schema once, use every layer.

Production-grade signals

The screening logic comes from a live XRPL trading pipeline: issuer blackhole checks, clawback hard-rejects, LP-burn verification. The blocklist contains rugs we caught with real money on the line.

Every check makes it smarter

Usage is metered, feedback is ranked, and each screened token or issuer feeds the threat data. More agents → better verdicts → more agents. Knowledge compounds.

The coverage gap

Token-security APIs (GoPlus, Webacy) stop at EVM chains and Solana. The XRP Ledger — with its own AMM and its own rug mechanics — had no equivalent. Now it does, built agent-native from day one.

Quickstart

Point any MCP client at https://cerberusindex.com/mcp and call a tool. Unpaid calls return x402 payment terms; x402-aware clients pay in USDC and retry automatically. Full walkthrough: Connect your agent.