Is this token safe to buy or accept? For an ERC-20 on Base or Ethereum: owner powers from its code (mint, blacklist, fees, pause, trading switch, limits), upgradeable proxy, renounced ownership, DEX liquidity (Uniswap v2/v3, Aerodrome vs WETH/USDC with USD estimate) and OFAC screening of token and owner. Verdict, score, reasons. $0.01 USDC per call.

Total value locked across DeFi, the top chains by TVL, and aggregate stablecoin market cap. Stablecoin supply is the dry powder of the whole market: rising supply with rising TVL is risk appetite expanding, falling supply is capital genuinely leaving crypto rather than rotating inside it.

Large ETH transfers (>100 ETH) to and from known exchange wallets in the last 2 hours. Inflows to exchanges are potential distribution pressure, outflows suggest accumulation or a move to cold storage — but read them as context, not instruction: a single large transfer is often internal rotation between wallets of the same desk.

Bitcoin network health: hashrate, difficulty, transaction count, estimated volume, miner revenue and the largest recent mempool transactions. A congested mempool carrying large transfers can front-run an exchange flow you will only see in price an hour later.

S&P 500, gold and Nasdaq 100 quotes as the traditional-market leg of a crypto read. BTC trades as a high-beta risk asset far more often than as digital gold, so equities direction is usually the better correlation to check before blaming a crypto-native cause. Priced as ETFs, so use direction and percent change, not the absolute level.

Ethereum address activity: ETH balance plus recent large transfers in and out. Point it at an exchange wallet, a treasury or a wallet a headline just named, and get the movement behind the story rather than the story.

AI crypto analyst, pay-per-call in USDC: ask any question on Bitcoin, Ethereum, altcoins or macro; the agent picks live market data tools (prices, funding rates, open interest, Fear & Greed, macro rates), runs them and returns one JSON object: regime, confidence, key_data with the numbers behind it, bull and bear scenarios, invalidation_level, summary. POST {question}.

BTC, ETH and USDT dominance as a share of total crypto market cap — the regime filter to check before sizing any altcoin position. Rising BTC dominance means capital rotating to relative safety; rising USDT dominance means capital leaving for stablecoins, a broader risk-off. Falling BTC dominance with rising ETH dominance is the classic rotation signal.

Scheduled macro catalysts ahead — FOMC, CPI, PCE, Non-Farm Payrolls, GDP, Retail Sales — from the official US release calendar. The dates that decide whether a position should be sized down before the print. Returns the scheduled release, not analyst consensus.

Crypto Fear & Greed Index: current and previous value with classification, plus recent history. A contrarian input — extreme fear has historically marked accumulation zones and extreme greed distribution zones — but it is a filter, never a trigger on its own. Pair it with funding to separate real capitulation from a crowded short.

Base token safety check: Claude Haiku 5.5 (Anthropic) writes a rug pull / honeypot verdict (SAFE/CAUTION/RISKY/DANGER) from on-chain data, a tagline, a short explanation. Deterministic fallback if Claude is slow, honestly labeled. GET /token-card/sample.

What the options market expects price to do, not what leverage is doing: the 30-day implied volatility index for BTC and ETH, with its percentile over the last month and the expected move in percent over 1, 7 and 30 days. Funding tells you who is crowded; this tells you how much movement is already priced in — and whether hedging is cheap or expensive relative to where it has been.

What a trade actually costs before it is placed: live order book depth on Bybit and OKX, walked at your size. Give it a notional in dollars and it returns the average fill price, slippage against mid in basis points and dollars, the spread, the depth within 1% of mid, and which venue is cheaper for that exact size — separately for buying and selling. A size the book cannot absorb is reported as unfilled, never priced by extrapolation.

The answer, not the inputs. One call returns a stance for the coin — risk-on, risk-off, fragile or neutral — built from funding and open interest percentiles, sentiment, options implied volatility, the macro regime and what it costs to execute at your size. Every driver comes back with its number and its reading, plus the conditions that would invalidate the call. The rules are deterministic: the same market gives the same verdict.

Funding rate and open interest for one coin, broken down venue by venue across Binance, Bybit, OKX and Hyperliquid, using each venue's largest-open-interest perpetual. Use it when venue divergence matters: one venue far more positive or negative than the rest is localised positioning, not market consensus — the setup that precedes a squeeze on that venue alone. Returns per-venue rows plus average funding, total open interest and the spread.

Perpetual funding rate and open interest for up to 10 coins in a single call, across Binance, Bybit, OKX and Hyperliquid. Answers 'where is the leverage right now?' for a whole watchlist at once. Thin venues are excluded: averaging every venue that publishes a rate overstates BTC funding by ~30x, because a near-empty book printing 9% per 8h counts the same as a deep one. Funding comes per 8h and annualized, so 0.0067% reads as 7.3% a year.

Use this service before importing or publishing JSON to check it against a restricted draft-07 schema. Send dataJson and schemaJson as JSON-encoded strings. Max 256 KiB UTF-8 each, depth 32; bounded node and work limits. Supports types, required properties, items, enums and numeric/length bounds. No regex, combinators or external references. Invalid business data returns valid=false.

Analyze raw HTML you provide and get structured SEO data as JSON: title, meta description, canonical, H1-H6 outline, links, JSON-LD blocks and structural checks (single H1, heading order, duplicates, JSON-LD validity). Use it to audit a page before publishing. No URL fetching, no JavaScript execution. Max 64 KiB of HTML.

Classify the search intent behind keywords: for each keyword, the dominant intent (informational, navigational, commercial or transactional) with a probability, plus any secondary intents. For mapping a keyword list to funnel stage and content type. Up to 1000 keywords per call.

Batch npm license triage for coding agents — pass packages from a lockfile (name@version list, max 20). Returns per-package SPDX + permissive-default compatible, rollup counts, and overall yes|no|unknown. Not a compliance SBOM. Use before merge or mass install; single-package use GET /license/check.