All projects matching “clean”

Japanese address normalization (住所正規化): standardize a JP address into clean prefecture / city / town / chome-banchi components. Fixes fullwidth/halfwidth digits, old-form kanji, and chome-banchi notation variants, and resolves the 7-digit postal code. Returns a 0-1 confidence score so agents can flag low-quality input. Use for address validation, cleansing, deduplication, and canonicalization befo

Google Patents search — find patents by technology, inventor or assignee with date, country, legal-status and classification filters; clean JSON with titles, snippets, dates, owners, PDFs and patent IDs.

Fetch any public URL and return its content as base64 with the content type and a ready-to-use data URI. Embed images, PDFs and files directly in LLM and API calls without downloading them yourself. Clean JSON.

Fetch a URL and extract clean main-content text with title, description, word count, and char count; boilerplate (nav/footer/sidebar/ads) is stripped.

Fetch any public web page or PDF and return clean Markdown, plus title, description, author, publish date, JSON-LD and outgoing links. Built for LLM agents.

Web search for AI agents: query in, ranked results out (title, URL, snippet) as clean JSON. Sovereign metasearch — no API key, no upstream account. Also does news search (type=news). Priced to undercut the incumbents.

Fetch any public URL and return a clean, readable markdown or JSON extract for AI agents. Pass the target as ?url=<encoded>. Ideal when you need to read webpages, docs, or articles without a browser. Blocks script/style tags, normalizes links and images, strips HTML.

eBay listings search — live prices, condition, shipping, seller reputation, auctions & Buy It Now across supported regional eBay sites, with category, price band, condition, seller & location filters, sorted, as clean JSON.

Encode text into international (ITU) Morse code. Letters are separated by spaces and words by "/". Supports A-Z, 0-9 and common punctuation; unsupported characters are clean 400s.

Test whether a RE2 pattern matches anywhere in the input. Returns { matched }. RE2 guarantees linear-time evaluation, so this is safe against catastrophic backtracking. An invalid pattern is a clean 400; no match is a successful 200.

Full transcript of a YouTube video from its URL — captions/subtitles as clean plain text in your language. YouTube transcript, video to text, captions extraction, subtitle download, speech to text, multilingual. 105 languages supported, including English, Spanish, Portuguese, French, German, Hindi, Arabic, Chinese, Japanese, Korean, Russian, Indonesian and Vietnamese. Pay per call, no account or A

Extract structured metadata from any URL: page title, meta description, Open Graph and Twitter Card tags, canonical URL, favicons, language, generator and robots directives, plus RSS/Atom feed links. Give it a URL, get back clean SEO and social-share metadata. No account, no API key — pay per call with x402.

Solve the linear system A·x = b for a square coefficient matrix A and vector b via Gaussian elimination with partial pivoting. A non-square A or a b of the wrong length is a clean 400. A singular system is NOT an error — it returns a successful { solvable:false, solution:null }.

Compute the standard English readability formulas — Flesch Reading Ease, Flesch–Kincaid grade, Gunning fog, SMOG, Coleman–Liau and ARI — along with the underlying sentence/word/syllable statistics. Empty text is a clean 400.

Trace an HTTP redirect chain to its final destination without downloading page bodies. Returns every hop (URL + status), the final URL, and safety warnings — link-safety grounding for agents before they visit or cite a URL. Clean JSON.

Email authentication posture check: parses SPF, DMARC and DKIM plus MX for any domain and grades deliverability/anti-spoofing gaps (SPF +all, SPF >10 lookups, DMARC p=none, missing DKIM). Clean JSON for deliverability and security-ops automation.

Full transcript of any YouTube video by video id or URL: captions/subtitles as clean plain text plus timestamped segments, the video's chapter markers, and every caption track it offers so an agent can pick a language and re-fetch. YouTube transcript, video to text, captions extraction, subtitle download, closed captions, speech to text, multilingual, asr and authored tracks. No API key.

Call before installing or updating an npm or PyPI dependency. Returns a verdict (ok, caution, avoid, unknown) plus evidence signals: install script added, maintainer or publisher change, very recent publish, build provenance dropped, known malware and vulnerability advisories. A clean result is not a guarantee of safety.

Fetch a URL and return clean, token-efficient markdown

Flight search by departure and arrival airports in the URL — compare live itineraries, fares, schedules, stops and emissions across airlines, similar to Google Flights, as clean JSON.

Extract raw artifacts from images: OCR text and detected objects. No interpretation, just clean machine-readable data. Powered by GPT-4o Vision. Payment via USDC on base.

Area (Heron's formula), perimeter and interior angles of a triangle from its three side lengths. Sides that violate the triangle inequality are rejected as a clean 400.

Official French company data (SIRENE/RNE), cleaned: identity, geolocated HQ, NACE code, headcount, latest revenue and net income.

Ordinary-least-squares linear regression of y on x: slope, intercept, correlation r and coefficient of determination r². Pass predictAt to also evaluate the fitted line at a given x. Rejected as a clean 400 if lengths differ or x has zero variance.

Scrape any public web page and get clean structured JSON (title, text, meta). Pay per call with USDC on Base.

URL-to-Markdown web extraction and context compression for AI agents. Fetch a public web page and return clean, task-relevant, citation-verifiable LLM context with exact excerpts, locators, hashes, and estimated tokens saved. Use before an agent summarizes, answers questions from, or cites a long or noisy URL. Deterministic and model-free. Public HTML only; no PDFs, logins, or JavaScript-only page

Markdown only: fetch a public web page (robots.txt respected) and return clean Markdown plus title/word count. Cheaper than /v1/extract when you do not need structured product/table JSON.

Extract readable content from any URL as clean Markdown. Strips navigation, ads, footers. Returns title, author, date, word count. 15-min cache per URL. Use when an agent needs webpage content as text.

Fetch a public URL with SSRF protection, clean extraction, and semantic prompt-injection analysis.

Convert a decimal string to an exact reduced fraction. For a repeating decimal, set repeat to the number of trailing fractional digits that repeat (e.g. value='0.1666', repeat=1 → 1/6; value='0.3', repeat=1 → 1/3). A non-numeric value is a clean 400.