information-retrieval

Check a specific dataset, paper, repository, or model URL against the indexed open-science corpus and receive a conservative, source-linked commercial-use and redistribution status. Never infers permission from an absent or ambiguous licence -- returns unknown rather than guessing.

Search the NCBI NLM Catalog journal/resource database and return matching UIDs for query {q} via E-utilities. Query: ?q= Returns JSON.

Colombia (datos.gov.co, SECOP procurement system): public procurement processes/contracts matching a free-text search. Query: ?q= Returns JSON.

SYNTHORA sanctions-screen: sanctions screening with MATCH / CLEAR verdict over 4 official downloads (OFAC SDN primary and alias lists, UN, OpenSanctions). Multilingual, tested on Cyrillic transliterations.

Search Yelp businesses by query and location.

View the roadmap

Classify text into one of the labels you provide, with a confidence score and a short justification. Useful for routing, triage, moderation queues and tagging pipelines where an agent needs a cheap deterministic decision per item.

text-metrics: Returns character count, word count, and average word length for any given string.

Returns summary statistics: total indicators, countries, data points, and last update timestamps.

Find source-linked scientific papers by topic. Returns metadata pointers, authors, dates, licence evidence when supplied by the source, and corpus provenance; it does not sell paper text or infer reuse rights.

Read robots.txt rules and sitemap declarations before crawling or indexing a domain.

Email validation — Syntax check + live MX/deliverability + disposable-domain flag.

DNS lookup — Resolve DNS records via DNS-over-HTTPS.

Extracts URLs from a specified domain.

Return title, description, OpenGraph card, canonical URL, feeds, favicon and detected technologies for one URL.

Screen a person/company/vessel name against OFAC (SDN + Consolidated), UN, EU, UK, Canada, Australia, France, Switzerland, US export-control (BIS/State), and US SAM procurement/benefit exclusions. Deterministic typo-tolerant fuzzy matching with versioned list snapshots.

Agent researches a topic (web) and returns a concise, structured, cited brief. ?topic=. Pay-per-call in USDC on Base.

Returns a synthesized, citation-backed answer to a research question, composed from verified internal sources with confidence scoring.

Embeds one query and up to 64 candidates locally, then returns cosine scores and stable descending ranks.

Returns letter details for a specified identifier with a 7-day time window.

Raw source verification for a TeleSint record. Pass id (UUID from any items[].id). Returns original defanged message text and source language alongside the AI summary for provenance checks.

CVE lookup

MD5 hex digest of text. Call when you match legacy content hashes (not for security). Returns hex digest for $0.001 USDC via x402 on Base. Legacy checksum only—not password hashing. No network, no keys, no retention. Results are advisory; the caller owns budgets, auth, and production controls.

Harvests the curated feed set for one topic and language, groups the headlines into events and returns one entry per event. The source count is the product: ten outlets on one story is a different signal than one. Never returns third party article bodies, only headline, own abstract and link.