Search US patent applications and grants via the USPTO Open Data Portal. Query: q (required, 2+ chars), yearFrom / yearTo (optional filing-year bounds), applicationType (optional: Utility|Design|Plant|Reissue), limit (1-100, default 10), offset (0-based, default 0). Returns { total, returned, offset, limit, hits[{ applicationNumber, title, applicationType, firstInventor, inventors[], applicants[], filingDate, effectiveFilingDate, status:{ code, description, updatedAt }, cpcSymbols[], uspcSymbol, url }] }. URLs link to USPTO Patent Center for the public file wrapper.

Sweep a person's name across five US public registries in one call: FINRA securities brokers, federal-court attorneys (CourtListener), federal inmates (BOP), Texas trade licenses (TDLR), and Texas real-estate licenses (TREC). Returns one block per registry with found/error status, match count, and the matching records — name-matched CANDIDATES, deliberately not merged into one identity (same name does not mean same person; verify with each registry's identifier before acting). Due-diligence, KYC screening triage, and background-research staple. Each registry is also a standalone endpoint (/api/license/broker, /api/law/attorney-lookup, /api/gov/inmate-locator, /api/license/trades, /api/license/real-estate) for follow-up by identifier.

Parse, validate and canonicalize a phone number using Google's libphonenumber metadata. Query: number (required; E.164 like "+14155552671" or national like "(415) 555-2671"), country (optional ISO 3166-1 alpha-2 like US/GB/JP — required when number is not E.164). Returns {input, valid (matches a real number range), possible (right shape/length), e164, national, international, rfc3966, country, countryCallingCode, type (mobile|fixed_line|fixed_line_or_mobile|toll_free|premium_rate|shared_cost|voip|personal_number|pager|uan|voicemail|unknown), possibleCountries[]}. Returns valid:false (not an error) for unparseable input.

Find points of interest near a coordinate. Backed by OpenStreetMap via the Overpass API (free, public, ODbL). Returns name, OSM id (e.g. node/123 — deep-linkable to openstreetmap.org), latitude, longitude, distance in meters, composed street address (when present), phone, website, opening hours, brand, and cuisine tags. Results sorted nearest-first. Supported categories: restaurant, cafe, bar, fast_food, gas_station, ev_charging, parking, atm, bank, hospital, pharmacy, clinic, doctor, dentist, police, fire_station, post_office, library, toilets, school, university, supermarket, convenience, hotel, hostel, museum, attraction, park, playground. Query: lat (-90..90), lon (-180..180), category (one of the supported names), radius_m (1-10000, default 1000), limit (1-100, default 20).

A wallet's Polymarket on-chain activity feed: trades, merges, splits, redeems, conversions and rewards, newest first. Pass the wallet (proxy) address; optionally filter by type (TRADE/MERGE/SPLIT/REDEEM/REWARD/CONVERSION) and cap with limit. Each entry returns type, side, outcome, share size, price, USD value, timestamp, the market title/slug, conditionId, and tx hash. Read-only from Polymarket's Data API.

Polymarket crypto up-or-down markets — the recurring 'Will <asset> be up or down by <time>?' series for BTC, ETH, SOL, SPX and more (hourly/daily). Each event returns its question, dates, USD volume, and the nested Up/Down markets with live outcome prices (implied probability the asset closes higher). Filter count with limit. Read-only mirror of Polymarket's up-or-down tag.

Browse Polymarket events (an event groups related markets — e.g. an election, a tournament, a price ladder). Optional keyword q, closed filter, order (volume/volume24hr/liquidity), and limit/offset. Each event returns its title, tags, USD volume + 24h volume, liquidity, dates, and the nested markets with their outcomes + live outcome prices (implied probabilities) and CLOB token ids. Read-only mirror of Polymarket's Gamma events.

Top holders of a Polymarket market, grouped by outcome token (conditionId). Each holder: wallet, trader name, position size, outcome index, and verified flag. Reveals concentration and smart-money positioning per outcome — the holder-side complement to predict.trades (flow) and predict.whales (large trades). Read-only from Polymarket's Data API.

Browse Kalshi events (an event groups related markets, e.g. an election or a game). Filter by status / series_ticker; page with limit + cursor. Returns event ticker, series, title, category, and market count. Read-only (no Kalshi key needed).

A single Kalshi market by ticker. Returns yes/no bid+ask and last price (dollars, implied probability), volume, liquidity, open interest, open/close times, and settled result. Read-only (no Kalshi key needed).

Browse Kalshi regulated prediction markets. Filter by status (open/closed/settled), event_ticker, series_ticker, or specific tickers; page with limit + cursor. Each market returns yes/no bid+ask and last price (in dollars, 0-1 = implied probability), 24h + total volume, liquidity, open interest, open/close times, and result if settled. Read-only mirror of Kalshi's market list (no Kalshi key needed).

Order book for a Kalshi market by ticker: resting yes and no bids with price (dollars) and size (contracts). Optional depth. Read-only (no Kalshi key needed).

Recent trades on Kalshi, optionally filtered to one market ticker. Returns each trade's ticker, contract count, taker side, yes/no price (dollars), and time. Page with limit + cursor. Read-only (no Kalshi key needed).

Polymarket smart-wallet leaderboard: top traders ranked by realized volume or profit (P&L). Choose the time window (1d/7d/30d/all) and rankBy (vol/pnl); cap with limit. Each entry returns rank, wallet (proxy) address, trader name, X handle, verified badge, total USD volume, and USD P&L — for tracking smart money. Pair the wallet with predict.positions / predict.activity / predict.wallet. Read-only from Polymarket's Data API.

Active prediction markets on Limitless Exchange (on-chain, Base; keyless). Each market: conditionId, title, slug, description, status, live YES/NO prices, volume, liquidity, collateral token, expiration, and categories/tags. A second venue alongside the Polymarket (predict.markets) and Kalshi (predict.kalshi-markets) clusters — for cross-venue prediction-market coverage.

A single Polymarket market by conditionId, slug, or id. Returns the question, outcomes + live prices (implied probabilities), CLOB token ids, USD volume + liquidity, dates, description, and URL. Pass one of conditionId / slug / id.

Browse Polymarket prediction markets. Filter by active/closed, order by volume/liquidity/endDate, page with limit/offset. Each market returns its question, outcomes + live outcome prices (implied probabilities), CLOB token ids (use with predict.price / predict.orderbook / predict.price-history), USD volume + liquidity, open/close dates, and the Polymarket URL. Read-only mirror of Polymarket's public market list.

Cross-venue equivalent-market pairs for arbitrage / relative-value spotting. We fuzzy-match open Polymarket vs Kalshi markets by title-token overlap and return candidate pairs with both venues' YES prices and the implied-probability spread (Polymarket YES minus Kalshi YES). Tune with minScore (0-1 title-similarity floor) and limit. Heuristic — a high score is a strong candidate, not a guaranteed identical contract. Read-only, keyless. No public source maps these directly; the pairing is derived by 2s.io.

Full CLOB order book (bids + asks with price and size) for a Polymarket outcome token. Use the clobTokenIds from predict.markets. Read-only depth snapshot from Polymarket's CLOB.

A wallet's open Polymarket positions. Pass the wallet (proxy) address; cap with limit. Each position returns the market title + outcome, share size, average vs current price, initial vs current USD value, unrealized USD P&L and percent P&L, whether it's redeemable, and the conditionId/slug. For inspecting any trader's live book (e.g. a wallet from predict.leaderboard or predict.whales). Read-only from Polymarket's Data API.

Live best bid, best ask, and midpoint for a Polymarket outcome token (CLOB token id — get it from predict.markets clobTokenIds). The midpoint is the market's implied probability for that outcome. Read-only from Polymarket's CLOB.

Time-series price (implied-probability) history for a Polymarket outcome token. Pass the CLOB token id and an interval (1h, 6h, 1d, 1w, 1m, max). Returns timestamped price points — for charting how a market's odds moved. Read-only from Polymarket's CLOB.

Cross-venue prediction-market keyword search across Polymarket, Kalshi, and Limitless Exchange in one call. Pass q (required); filter by status (open/closed) and cap per-venue with limit. Each hit is tagged with its venue and returns the question, live YES/NO price (0-1 implied probability), USD volume, close date, and the venue URL — for finding the same real-world question wherever it trades. Read-only, keyless.

Sports prediction markets across Polymarket and Kalshi in one call — game lines, futures, and props (NFL/NBA/MLB/NHL/soccer/tennis/etc). Optionally narrow with league (a tag/category keyword, e.g. nfl, soccer, world-cup) and cap with limit. Each hit is tagged with its venue and returns the question, live YES/NO price (implied probability), USD volume, close date, and the venue URL. Read-only, keyless.

Recent Polymarket trades. Filter by market (conditionId) and/or user (wallet address); page with limit (max 500). Each trade returns wallet, trader name, side (buy/sell), outcome, size, price, USD notional, timestamp, market title, and tx hash. Read-only from Polymarket's Data API.

A Polymarket trader's portfolio by wallet address: total portfolio USD value plus open positions (market title, outcome, size, average vs current price, current value, and unrealized PnL). For tracking a wallet's prediction-market book. Read-only from Polymarket's Data API.

Polymarket whale radar: the largest recent trades by USD notional across all markets, ranked. Optional minUsd floor and limit. Each entry includes wallet, trader name, market, side, outcome, size, price, USD value, and tx hash — for tracking smart-money / large positioning. Read-only from Polymarket's Data API.

Decode and validate a product barcode (GTIN / UPC-A / EAN-13 / EAN-8 / GTIN-14 / ISBN-13 / ISSN). Returns the GS1 mod-10 check-digit validation (valid flag + expected digit — an LLM can't reliably compute this), the symbology, the zero-padded GTIN-14 canonical form, and the GS1 prefix decoded to the issuing GS1 member organisation and its country (NOTE: that's the country of the org that issued the prefix, not necessarily where the item was made). Flags restricted-distribution (in-store/variable-measure) prefixes and Bookland (ISBN) / serial (ISSN) ranges. Also attempts a FRESH best-effort product-identity lookup (name, brand, category, image) across the open Open Food/Beauty/Products/Pet Food Facts federation (ODbL) — reliable for food, beverage, and beauty items, sparse for general retail (electronics/apparel), and may be null. Always live (no caching). Set identity=false to skip the lookup and decode only.

NYC deed + mortgage history for a tax lot via ACRIS (Automated City Register Information System) legals dataset, keyed by BBL. Each row carries a unique documentId you can use to drill into the ACRIS master dataset (the master URL pattern is included in the response) for full details: parties, consideration amount, document type (DEED, MORTGAGE, ASSIGNMENT OF MORTGAGE, etc.). Use `property.nyc-parcel-lookup` first to convert an address to a BBL. Returns most-recent records first.

NYC tax-lot lookup via PLUTO (Primary Land Use Tax Lot Output) — every tax lot in the city with owner, zoning, lot/building area, year built, classification, lat/lon, community + school + council + police districts. Pass `bbl` (10-digit Borough-Block-Lot composite, e.g. 1010110001 for the Empire State Building) for an exact lookup, OR pass `address` (partial-match) optionally constrained by `borough` (name or 2-letter code MN/BX/BK/QN/SI). The BBL returned here is the universal join key for all other property.nyc-* endpoints — fetch parcels first, then chain into deed-history, permits, or violations.

NYC Department of Buildings permit issuance — every construction, alteration, or demolition permit ever issued. Search by `bbl` (10-digit) or `address` (street_name substring). Filter by jobType (A1=Major Alteration, A2=Minor Alteration, A3=Minor Cosmetic, NB=New Building, DM=Demolition, etc.) or permitStatus (ISSUED, IN PROCESS, RE-ISSUED, REVOKED, etc.). Each row carries job + permit numbers, work type, building type, residential flag, filing/issuance/expiration dates, estimated fee. Use for construction-history agents, code-enforcement research, and zoning compliance checks.

NYC Housing Preservation & Development (HPD) violations — every notice of violation issued at a multi-family building. Search by `bbl` (10-digit) or `address` (street_name substring). Filter by `classCode` (A=least severe, B=hazardous, C=immediately hazardous) and `currentStatusOnly=true` to limit to open violations. Each row carries violation id, building id, full address + apartment + story, inspection + approved + certify-by + correct-by dates, current status + status date, and the narrative NOV description. Use for landlord-history, code-enforcement, and tenant-rights agents.

Recent earthquakes near a coordinate. Returns each quake with magnitude, place name, time (ISO), latitude/longitude/depth, tsunami flag, USGS event URL, and distance-from-query in km — sorted by time descending. Real-time data, post-LLM-training-cutoff. Backed by USGS FDSN event API (public domain). Query: lat (-90..90), lon (-180..180), radius_km (1-1000, default 500), hours (1-720, default 24), min_magnitude (0-10, default 2.0).

Search the Recreation Information Database (RIDB) — the single source of truth for federal recreation lands and programs across NPS, USFS, BLM, USACE, BOR, FWS, NARA. Pick a resource: recareas (designated recreation areas), facilities (individual sites — campgrounds, day-use, visitor centers), campsites (individual reservable campsites with attributes), permits (special-use permits, lotteries), tours, events, activities (taxonomy lookup). Filter by free-text query, state, activity ID, or lat/lon + radius (miles, max 50). Used by every federal-recreation-related agent: trip planning, permit chasers, fishing/hunting locator, campground availability tools.

Look up an npm package by name (supports scoped packages like @sentry/node). Returns description, homepage, repository, license, author + maintainers list, keywords, distTags (latest/beta/rc/etc.), latest version + publication date, and the 50 most recent versions with their publish dates + deprecation status. ~3M JavaScript/TypeScript packages. npm public registry.

Look up a Python package on PyPI by name. Returns version, summary + long description, content-type, project URLs (Homepage, Source, Bug Tracker, Docs, etc.), license, author + maintainer (with emails), keywords, classifiers (top 30), requires-python spec, requires-dist (top 100 runtime deps), the 50 most recent releases with publish dates, and any yanked versions in that window. PyPI public registry.

Resolve a researcher by ORCID iD. Pass orcid (e.g. 0000-0002-1825-0097, with or without the https://orcid.org/ prefix). Returns the researcher name, affiliations (employments and educations with organization, role, department, years, and a current flag), a total works count, and a works summary (title, type, year, DOI when present). Data: ORCID Public API, free and keyless; per-record visibility is set by the record holder. The canonical author key in scholarly metadata — pair with /api/papers/search, /api/paper/doi-lookup, and /api/research/org (ROR institutions). Use to verify an author identity, current institution, or publication record before citing.

Search US federal biomedical research grants via NIH RePORTER. Filter by term (free-text over project title, terms, and abstract), org (funded institution name), pi (principal-investigator name), and/or fiscalYear; page with limit (1-50) and offset. Returns each award with project number, title, fiscal year, award amount (USD), contact PI, organization (name, city, state, country, department), funding agency/institute, and project start/end dates — newest fiscal year first, with a total match count. Data: NIH RePORTER, free and public-domain. Fresh post-training award data agents cannot know; pair with /api/research/org (ROR) and /api/research/author (ORCID). Use for funding landscape research, competitor/lab tracking, and grant-opportunity context.

Resolve a research organization via the Research Organization Registry (ROR). Pass id (a ROR id like 056y0v115 or its full URL) to fetch one org, or name to search by free text. Returns the canonical ROR id and name, organization type, status, founding year, location (city, country, GeoNames coordinates), website, Wikipedia link, external identifiers (GRID, ISNI, Wikidata, Crossref Funder), parent/child relationships, and aliases/acronyms. Data: ROR, free and CC0. The canonical institution key in scholarly metadata — pair with /api/papers/search and /api/paper/doi-lookup (author affiliations) and /api/research/author (ORCID). Use to disambiguate an institution name to a stable id before joining datasets.

AI web search optimized for agents. Returns ranked results with the relevant extracted content of each page (not just a link + blurb), plus a relevance score. topic=news for recent reporting. Distinct from search.web (raw SERP) — this returns clean, LLM-ready page content per result.

Find the right 2s endpoint for a task using natural language. Pass q (e.g. 'check if a domain can be spoofed', 'is this CVE being exploited', 'screen a company for sanctions', 'decode a VIN') and get back the catalog's most relevant endpoints, ranked, each with its id, API path, method, group, description, and price. Deterministic keyword + synonym matching over the live endpoint registry (no LLM, no external calls) — a self-routing index so an agent (or its planner) can discover which paid call answers a question instead of reading the whole 290+ endpoint catalog. Returns the path you then call directly.

Live web search. Returns ranked results with title, URL, snippet description, site name, and page age — fresh information past any model training cutoff. Supports count (1-20), offset for paging, country (2-letter, e.g. US), freshness (pd=past day, pw=past week, pm=past month, py=past year, or a YYYY-MM-DDtoYYYY-MM-DD range), and safesearch (off/moderate/strict). Use for current events, fact verification, finding documentation, and research. Independent search index. For news-specific results with sources and timestamps see /api/news/search.

Authoritative MITRE ATT&CK (Enterprise) technique lookup. Pass id (e.g. T1059 or sub-technique T1059.001) for the canonical technique — name, tactics (kill-chain phases), description, platforms, sub-technique flag + parent, mitigations, and detection guidance — or query for a keyword search returning ranked techniques. Bundled current ATT&CK matrix (~700 techniques), zero external calls. Agents cite T-numbers and tactic names that must be exact; this returns version-pinned, citeable data instead of hallucinated IDs. For threat modeling, detection engineering, and report enrichment.

Authoritative MITRE CAPEC (Common Attack Pattern Enumeration) lookup. Pass id (e.g. CAPEC-66, or just 66) for the canonical attack pattern — name, abstraction, description, typical likelihood + severity, mapped CWE weaknesses (with names), and related patterns (with names) — or query for a keyword search. Bundled catalog (~615 patterns), zero external calls. The attacker's-eye complement to security.cwe (the defender's weakness view) — the CAPEC↔CWE cross-links let an agent pivot between how an attack works and the weakness it exploits, with exact citeable IDs.

Look up a CVE by id (e.g. CVE-2021-44228) across three authoritative vulnerability feeds in one call. Query: cve (CVE-YYYY-NNNN). Returns the canonical record — description, CVSS base score + severity + vector, CWE weakness ids, published/modified dates, reference links — plus whether it is on the US CISA Known Exploited Vulnerabilities catalog (with remediation due date and known-ransomware flag) and its EPSS exploit-probability score and percentile. The exploited and EPSS sections report independently, so one feed being unavailable does not fail the call. 404 if the CVE id is unknown. For triage, prioritization, and anti-hallucination on vulnerability claims.

CVE change feed — the CVE records MODIFIED within a time window, so an agent can incrementally maintain a vulnerability view instead of re-scanning. Pass since (YYYY-MM-DD or ISO datetime); until defaults to now (window must be ≤ 120 days, the NVD limit). Optionally narrow by keyword (product/text) or cpe (exact CPE). Returns each changed CVE with its id, published + lastModified timestamps, current vulnStatus (e.g. Modified, Analyzed, Rejected), best-available CVSS score/severity, description, and kevListed — whether it is now on the CISA Known-Exploited Vulnerabilities catalog (the high-signal flag for a poller). Newest modification first. Sourced live from NVD (NIST) + CISA KEV, free/keyless. Pair with security.cve for full per-CVE detail.

Find vulnerabilities affecting a product by searching the NIST National Vulnerability Database. Pass product (free-text keyword, e.g. "apache log4j", "openssl", "wordpress plugin contact-form-7") or cpe (an exact CPE 2.3 name, e.g. cpe:2.3:a:apache:log4j:2.14.1:*:*:*:*:*:*:*). Returns matching CVEs newest-first, each with its id, description, CVSS base score/severity/vector, and dates. Optional limit (1–50). For "what CVEs affect X" / surveying a product's vulnerability history — distinct from security.cve, which resolves a single CVE id across NVD + CISA KEV + EPSS. Free, keyless.

Authoritative MITRE CWE (Common Weakness Enumeration) lookup. Pass id (e.g. CWE-79, or just 79) for the canonical weakness — name, abstraction, description, extended description, ChildOf/ParentOf relationships (with names), and mapped CAPEC attack patterns (with names) — or query for a keyword search returning ranked matches. Bundled catalog (~970 weaknesses), zero external calls. Agents hallucinate CWE IDs and names constantly; this returns exact, citeable, version-pinned data. Pairs with security.cve (which returns CWE ids) and security.capec.

Does public exploit code exist for a CVE, and where? Pass cve (e.g. CVE-2021-44228). Returns hasPublicExploit, the count, hasMetasploitModule and hasVerifiedExploit flags, and the Exploit-DB entries (id, description, type, platform, date, verified, Metasploit flag, link). Bundled inverted index from the Exploit-DB archive (~25k CVEs). This is the triage signal BEYOND security.cve's KEV (exploited in the wild) + EPSS (exploit probability): is the vulnerability actually weaponized with available code? Use the trio together to decide how urgently to patch. Absence is not proof no exploit exists (private/other archives not covered).

Fetch a URL and grade its HTTP security headers. Pass url (scheme optional — defaults to https). Returns an overall letter grade + score, the list of present/missing headers, and a per-header analysis with the live value and specific issues for: Strict-Transport-Security (HSTS max-age/includeSubDomains), Content-Security-Policy (flags 'unsafe-inline'/'unsafe-eval'/missing default-src), X-Frame-Options or CSP frame-ancestors (clickjacking), X-Content-Type-Options (nosniff), Referrer-Policy, Permissions-Policy, and Cross-Origin-Opener/Resource-Policy. Also flags Server/X-Powered-By info disclosure. Analyzed from the target's LIVE response headers through an SSRF-guarded fetch (private/loopback targets refused) — an LLM cannot see a site's current headers. For web-app security review, vendor assessment, and CI gates.