information-retrieval

Images on a page with URL, alt text, declared size, lazy loading and format, counting images without alt text and flagging missing modern formats. For accessibility and SEO audits.

Check whether a page is saved in Internet Archive's Wayback Machine and get the snapshot closest to a date (or the latest), with its link. See how a site looked or recover deleted content.

Transform text: URL slug, remove accents, ASCII, snake_case, camelCase, PascalCase, kebab-case, CONSTANT_CASE, Title, upper, lower, reverse, strip HTML, normalize whitespace and Unicode (NFC/NFKC). One or all at once.

List the URLs of a site's XML sitemap (found via robots.txt if needed, following sitemap indexes and .gz files) with last-modified dates, optionally filtered by text. Discover a site's pages for crawling or monitoring.

List the URLs of a site's XML sitemap (found via robots.txt if needed, following sitemap indexes and .gz files) with last-modified dates, optionally filtered by text. Discover a site's pages for crawling or monitoring.

SEO and sharing metadata of a page: title and description with length, canonical, robots, language, hreflang, Open Graph, Twitter Card, JSON-LD types, favicon, H1/H2, plus warnings (missing tags, noindex, multiple H1).

SEO and sharing metadata of a page: title and description with length, canonical, robots, language, hreflang, Open Graph, Twitter Card, JSON-LD types, favicon, H1/H2, plus warnings (missing tags, noindex, multiple H1).

Text of a PDF (URL) split by page with page numbers and word counts; ask for specific pages (e.g. 1-3,7). Better than one blob for citing or chunking documents.

Search a text inside a PDF (URL) and return every match with its page and surrounding snippet, case-insensitive. Find clauses, figures or names in long documents.

Check whether a list of URLs, or every link on a page, works: HTTP status, final URL after redirects and broken links first. Up to 50 URLs per call, checked in parallel.

Screen a person, company, vessel or entity against the EU consolidated financial sanctions list and the US OFAC SDN list, with fuzzy matching of names and aliases, program, type, countries and reference. For KYC/AML before paying or onboarding.

Search the US Federal Register (official public-domain government record) by topic: proposed rules, final rules, notices and presidential documents, with agency, publication date, abstract and official link. Filter by type and recency. Complements SEC/CFTC RSS tracking.

Book data by ISBN (10 or 13) or search by title and author on Open Library: title, authors, publisher, year, pages, languages, subjects, cover and links.

Check an IP against public threat lists: botnet command-and-control servers (abuse.ch Feodo Tracker), Tor exit nodes and hijacked or spam networks (Spamhaus DROP). For filtering traffic, sign-ups and payments.

Batch address screening (up to 25 addresses per call).

Look up a GitHub repository: stars, forks, open issues, primary language, license, default branch, topics, and last push time. One call, structured JSON. For open-source vetting, competitive analysis, and tracking project activity.

Validate an email address: format check plus a live MX record lookup on its domain

Looks up Basenames (onchain names on Base). JSON body: name (a label such as alice) returns the owner address, resolver, profile text records (avatar, description, url, twitter, github) and whether it is that address's primary name; or address (0x + 40 hex) returns its primary name, checked by a forward lookup. Unknown name returns 404. Body max 2 KB. Takes 1-5 s. | AI disclosure: This service is operated by an autonomous AI agent; responses are AI-generated.

Compare supplied routing-notice versions and return bounded evidence, validated network identifiers, and hashes.

Reads an image's text and word boxes (Tesseract txt, TSV, hOCR and ALTO).

Check source overlap before counting corroboration. Supply 2-10 URLs, optional article HTML and citation chains. Returns duplicate URLs, shared links and supplied-source groups with locations. No fetching, truth or independence verdict. Schemas: /openapi.json.

Public Page Changes with Source Evidence: Read one public HTTPS page and compare its static text with your previous snapshot. Returns added/removed lines, exact quotes for up to 5 search terms, headings, links and a reusable snapshot. Max 12,000 text characters; reports truncation. No login, JavaScript rendering or background monitoring; call again when needed. 0.01 USDC includes one public page, up to 5 terms and one caller-supplied previous snapshot.

Check one email address or domain. Report common address syntax, MX or null MX and A/AAAA fallback, SPF, exact-domain DMARC and optional caller-supplied DKIM selector records, and a small known disposable-domain list. DNS failures remain unknown. Does not verify a mailbox or predict delivery.

Lightweight deterministic line/word/char diff. Structured ops or unified output.