RugSense

scored launch intelligence · live on Base · paid per call over x402

Every freshly-launched token, scored for safety and momentum into one AVOID / WATCH / HOT decision — so you (or your agent) skip the rugs and catch the real ones, in a single call.

Not a raw “new pairs” feed. A scored decision: honeypot & tax simulation, contract checks, holder concentration, and LP burn/lock (v2 & v3) — each launch ships a transparent per-signal checks[] and a safetyConfidence.

HOT

WATCH

AVOID

Latest scored launches — quick view; full onchain safety (honeypot/holder/LP + per-signal checks) ships with the API

Token	Tier	Score	Safety	Mom.	Liq.	Age	Flags
BASE 365Base 365	AVOID	18	25	11	$1,517	7m	LOW_LIQUIDITY
stETHLiquid staked Ethe	AVOID	18	25	11	$20	9m	LOW_LIQUIDITY
WEB3Web3	AVOID	18	25	11	$1,516	13m	LOW_LIQUIDITY
BASE 365Base 365	AVOID	10	15	5	$1,516	1m	LOW_LIQUIDITY
BASE 365Base 365	AVOID	10	15	5	$1,518	4m	LOW_LIQUIDITY

For agents & developers

One x402 call returns the scored, ranked list. No API keys, no signup — your agent's wallet pays $0.03 USDC per call on Base.

GET /api/launches/latest?tier=HOT&minSafety=60   → ranked scored launches
GET /api/token/{address}                         → score one token ("is it safe?")
402 Payment Required (x402 v2) → pay USDC → data

Discoverable on the x402 Bazaar / Agentic.Market. Drop-in MCP server gives any MCP client a get_base_launches tool. Specs: /openapi.json, docs/INTEGRATE.md.

How the score works

Deterministic, no LLM. Every launch runs through eight signals; a single dangerous one (honeypot, pullable LP, mintable) caps the composite to AVOID no matter how strong the momentum.

Honeypot + tax

buy/sell trade simulation

Source verified

contract is open-source

Proxy / upgradeable

can the logic change?

Mint / blacklist / pause

dangerous functions in bytecode

Ownership

renounced, or owner can act?

Holder concentration

top-10 % + whale count

LP burn / lock

can liquidity be pulled? (v2 + v3)

Momentum

liquidity, volume rotation, buyers

Each launch returns the per-signal checks[] (pass/warn/fail/unknown) plus a safetyConfidence, so you see why — not just a number.

Scores are a risk filter, not a guarantee — DYOR. Not financial advice.