Tool Catalog

Payment wires: every paid endpoint accepts x402 and MPP (Machine Payments Protocol) on the same 402 - see Paying with x402 and Paying with MPP. Agent402 is the applied layer of Agentic Finance: agents that pay and get paid on their own.

500+ endpoints + 70+ multi-tool Skill Packs. The utility tools are deterministic - no LLM in that serving path: same input, same output, full input/output schemas; the /v1 model gateway and the report products are model-backed and priced as such. Discover them machine-readably (don't hardcode this page):

• /api/find?q={task} - resolve a plain-language task to the right tool (route, price, schema, ready example) in one call, so an agent skips the token-heavy "search to find a tool" step. Also the catalog.find MCP tool on the connector.
• /api/pricing - slug, route, price, category, description for everything
• /openapi.json - OpenAPI 3.1 with schemas
• /tools - human-readable docs, one page per tool with a working example
• x402 Bazaar discovery extension - every 402 response self-describes

The headline tools (wallet-only)

These exist because an agent mid-task cannot give itself a browser, a paid search index, or a disk:

Tool	Price	What it does
search	$0.02	Live web search over a paid index, no signup: the wallet is the credential
answer	$0.08	Web answer with inline citations: a one-call "ask the web" the model couldn't reach otherwise
render	$0.02	Real headless Chromium, JavaScript executed - reads SPAs that extract can't
screenshot	$0.015	PNG of any public page (viewport or full-page)
extract	$0.01	Main-article extraction → clean markdown (title, byline, word count)
pdf-info, pdf-rotate, pdf-extract-pages, pdf-merge, images-to-pdf, pdf, pdf-to-markdown	$0.002–$0.01	Read and manipulate PDFs
media-info, audio-convert, audio-normalize	$0.005–$0.02	Real ffmpeg: probe, transcode to mp3, EBU R128 loudness normalize
memory-* (11 tools)	$0.001–$0.003	Durable wallet-keyed state + cross-wallet coordination. See Memory and Coordination
x402-quote, x402-verify, usdc-balance, tx-status, gas-estimate, transfer-authorization, ens-resolve	$0.002–$0.004	Non-custodial x402 payment toolkit - decode 402 quotes, verify settlements, read balances/gas/tx, build EIP-3009 authorizations, resolve ENS. Multi-chain: Base, Polygon, Arbitrum, Optimism, Ethereum. See Payments and x402
dns, meta, robots-check, email-validate, ip-info, http-check, tls-cert, sitemap, whois	$0.001–$0.005	Network truth: metadata, DNS, TLS, liveness
block-number, chain-info, block-info, contract-code, erc721-owner, event-logs	$0.001–$0.003	Chain reads: head block and chain metadata, a block by number/hash, deployed bytecode at an address, ERC-721 ownerOf, and decoded eth_getLogs event queries. Read primitives an agent can call without running a node or holding an RPC key
openapi-diff, openapi-lint, openapi-extract, openapi-to-curl, openapi-mock-response, openapi-search, openapi-validate-payload, openapi-redact, openapi-resolve-refs, openapi-security-summary, openapi-required-params	$0.002	API-kit - work an OpenAPI 3.x / Swagger 2.x spec end-to-end: find the right operation, see its effective auth, know the minimum inputs, build a runnable curl, mock a response, validate a payload, diff two versions, score agent-readiness, shrink for LLM context, inline $refs
fx-rate, gov-data, weather-forecast, weather-alerts, earthquakes, barcode-lookup	$0.003–$0.005	Live keyless data: ECB currency rates, data.gov datasets, NWS weather, USGS quakes, product barcode lookup
stock-quote, stock-dividends, stock-history, earnings-calendar	$0.003–$0.015	finance-kit: price, dividend history, OHLC history, and the upcoming/recent earnings calendar for any ticker. Fresh, no API key required
crypto-trending, crypto-global, crypto-price, crypto-market, crypto-history	$0.008–$0.015	crypto-kit: prices, market data, OHLC history, trending coins, total market cap. Multi-coin in one call
treasury-*, fred-*, cpi-yoy, fed-funds, sahm-rule, yield-curve-spread, world-bank-*	$0.005–$0.025	macro-kit: official macro time-series from the St. Louis Fed (FRED v1 + v2 bulk release observations), the US Treasury, and the World Bank
edgar-company-lookup, edgar-filings, edgar-company-concept, edgar-company-facts, edgar-xbrl-frame, edgar-insider-trades, edgar-13f-holdings, edgar-recent-ipos, edgar-search	$0.005–$0.025	edgar-kit, SEC EDGAR: ticker→CIK, filings, XBRL company facts and concepts, cross-company XBRL frames, insider Form 4, 13F holdings, recent IPOs, full-text search
image-resize, image-convert, image-thumbnail, barcode-decode	$0.003–$0.005	Pure-CPU image transforms + barcode/QR decode (jimp / zxing)

Crypto derivatives, DeFi and Solana intel

Market structure an agent can read directly, keyless and deterministic, at per-call prices. Every one is on the same x402 / MPP 402 as the rest of the catalog.

Tools	Price	What you get
perp-markets, perp-funding, perp-funding-screener, perp-basis, perp-open-interest, perp-klines, perp-orderbook	$0.002 to $0.003	Live perpetuals: every listed market with mark and oracle price, funding now and hourly history, the funding screener across the whole book, premium and predicted funding, open interest, OHLCV candles and order-book depth
options-summary, crypto-options-chain, options-ticker, options-volume	$0.002 to $0.005	The options book: index price and the implied-volatility index, call/put open interest and put-call ratio, a full chain by expiry with mark IV, per-instrument greeks (delta, gamma, vega, theta, rho), and onchain options protocols ranked by notional volume
defi-yields, defi-yield-history, defi-protocols, defi-protocol, defi-chains, defi-chain-tvl-history, defi-fees, defi-dex-volume, stablecoins, stablecoin-supply-history	$0.002 to $0.003	DeFi: screen yield pools by chain, project, token and TVL (stablecoin-only filter included), rank protocols and chains by TVL, read fees and revenue, DEX volume, and stablecoin supply with peg deviation, each with a history sibling
sol-token-safety, sol-token-report, sol-token-holders, sol-token-pairs, sol-token-search, sol-trending, sol-price, sol-swap-quote, sol-token-lookup	$0.002 to $0.010	Solana token due diligence: a graded safety check (mint and freeze authority, liquidity, holder concentration), the full risk write-up, holders, pairs, trending mints, live prices and a swap quote with the route
crypto-news, crypto-indicators, crypto-market-pulse	$0.004 to $0.005	Market context: headlines from major outlets normalized and deduplicated, technical indicators computed here from candles (RSI, MACD, EMA, SMA, Bollinger, ATR, VWAP), and a one-call breadth snapshot across every listed perp
coin-price-by-contract, coin-profile, coin-history, coin-ohlc, coin-market-chart-range, coin-categories, global-defi, exchanges, exchange-tickers, exchange-rates, coin-search, coins-list	$0.005 to $0.008	Broad market coverage, including price by token contract address on a named chain when an agent holds an address and no coin id
asset-transfers, token-balances, token-allowance, tx-receipt, block-receipts, token-price-history	$0.002 to $0.005	Indexed chain reads across the major EVM chains: a filtered transfer log, balances for named contracts, an ERC-20 allowance, a decoded receipt, a whole block summarised, and historical token prices
fc-cast-search, fc-channel-feed, fc-trending, fc-user-casts, fc-cast, fc-cast-replies, fc-channel, fc-user-search, fc-cast-metrics	$0.003 to $0.005	Farcaster: search casts, read channel and user feeds, pull a thread, and measure engagement on the onchain social graph
site-map, site-crawl	$0.005, $0.02	Whole-site structure on demand: enumerate a site's URLs from robots.txt, its sitemaps and internal links, then crawl breadth-first to clean markdown under hard page, depth and time budgets

Images and video (flat per call)

Generation priced per picture or per clip rather than per token, on the OpenAI wire, so the cost of a call is known before it is made. Point any OpenAI SDK at base_url https://agent402.tools/v1.

Endpoint	Price	What you get
POST /v1/images/fast	$0.02	Budget text-to-image, about two seconds a picture, inline base64 out
POST /v1/images/pro	$0.05	Higher-fidelity text-to-image, one picture a call
POST /v1/images/generations	$0.08	The flagship image route
POST /v1/videos/generations	$0.20	One silent 4-second 720p clip, MP4 inline base64, 16:9 or 9:16. A failed or timed-out generation is never charged

Report products (outcome-priced, same 402)

When the job is a whole report rather than a call, the /v1 report routes sell one finished, cited report per payment - POST /v1/research ($0.60, pro $0.85, max $1.10), /v1/research/market-brief ($0.85), /v1/dossier ($0.85, max $1.10), /v1/ticker-pack ($2.00), /v1/fund ($0.60, max $0.85), /v1/filing-report ($0.85), /v1/domain-audit ($0.60, pro $0.85), /v1/recall-report ($0.60), /v1/insider-report ($0.60), /v1/token-brief ($0.60), /v1/token-risk ($0.60, pro $0.85), /v1/linkedin-article ($1.10, a publish-ready article package with images at LinkedIn's sizes), and the deterministic /v1/ipo-report digest ($0.05). Same x402 / MPP 402 as every other route, plus prepaid card credits; people buy the same reports by card at /reports for $2 to $5, a price that includes payment processing. Inputs, cadences and the $5/month monitors are on Reports, Monitors and Credits.

The long tail (pure-CPU, also payable with compute)

150+ utilities at mostly $0.001: hashing/HMAC, base58/base32/base64, JWT decode+verify, UUIDs, CRC32, morse, HTML entities, token-count (exact OpenAI BPE), text-chunk (RAG), json-validate (JSON Schema), jsonl, text stats/dedupe/sort/truncate/diff (Levenshtein), JSON/CSV/YAML conversion and querying, date math and cron calculators, validators (email syntax, IP, IBAN-style checksums…), math/stats, QR codes, and unit-convert - one parametric endpoint (POST /api/unit-convert with {value, from, to}) covering every unit pair across length, mass, volume, area, speed, time, data, pressure, energy, power, angle, frequency, temperature.

Why would an agent pay $0.001 instead of writing the code? Because writing, testing, and debugging a CSV parser mid-task burns 10–100× that in tokens - and some sandboxes can't execute code at all.

SQL execution certificates (sql-guard + sql-cert-verify)

Two tools that make "an agent is about to run this SQL against production" a checkable gate rather than a hope:

Tool	Price	What it does
POST /api/sql-guard	$0.004	Reviews one SQL statement against a fixed, published risk catalogue and returns a verdict (pass / warn / block) with the risks named (unbounded UPDATE/DELETE, tautological WHERE, DROP, TRUNCATE, DROP COLUMN, statement stacking, COPY … FROM PROGRAM, GRANT/role changes, session_replication_role and trigger/constraint bypass, writes to pg_catalog). On pass it also returns an Ed25519 certificate binding that verdict to the SHA-256 of the exact statement.
POST /api/sql-cert-verify	$0.001	The gate your database layer calls before it obeys the agent. Checks the signature, the certificate version, the expiry, and that the statement's SHA-256 matches the one certified. Returns {valid, reason, payload} and never throws on a malformed token, so the executor always gets one uniform answer. Pass publicKey to verify a certificate issued by another deployment.

Three design points worth knowing:

• The certificate binds a hash, not a statement. An edited statement, even by one character, no longer matches, so a certificate can't be re-pointed at different SQL.
• Literals and comments are scrubbed before analysis, so the word DROP inside a string is never a false alarm.
• Signing is env-gated. SQL_CERT_SIGNING_KEY (a PKCS8 PEM) is the signing identity. Unset, sql-guard still returns full verdicts and says plainly that it cannot certify. It never returns an unsigned object shaped like a certificate. Rotating the key invalidates outstanding certificates, which live 5 minutes by default.

Honest scope: this is a lexical guard over a published catalogue. It catches the shapes that destroy production data. It is not a SQL parser and cannot know that a WHERE clause names the wrong tenant.

Quality guarantees

• Every endpoint is re-tested against its own documented example in CI before any deploy reaches production.
• Tools that can't be served honestly get removed rather than left to take money and 502 (this has happened - see Operations).
• Errors are structured: a specific message naming the invalid field, never a bare 500.