Nero Labs Agent Tools

Fetches one public web page and returns its main article text, cleaned of menus, ads and footers (Mozilla Readability), plus title, author, publish and update dates, site name, language, image and canonical URL from the publisher's own data. Honours robots.txt. Use it to read a link before summarising or quoting it. Charged only when text comes back; blocked, missing or non-HTML pages are free.

Checks one email address and returns valid, risky or invalid with the reasons: address format, whether the domain can receive mail (live MX lookup), known throwaway domains (mailinator and others, invalid), shared inboxes like info@ or sales@ (risky), likely typos such as gmial.com with the fix in didYouMean, and free providers such as gmail.com (flag only). Does not ping the mailbox (no SMTP), so

Merges two or more lists of JSON rows (for example leads from Google Maps, a CRM export and a spreadsheet) into one list and removes the duplicates between them by a key such as phone, email or website, ignoring spacing, case and http/www. Keeps the most complete row, fills its blanks from the duplicates and labels which lists each record was in. Up to 500 rows in all. $0.001 per row sent, $0.002

Analyses up to 50 customer reviews with AI: sentiment, star-rating match, complaint and praise themes, urgency, and optionally a ready-to-post owner reply each, plus a business report with top issues. $0.002 per review, $0.005 per reply drafted, $0.03 per business report (output reviews_only skips it). Runs nerolabs/review-analyzer.

Applies one plain-English AI instruction to every row of a JSON list (up to 50 rows), such as classify, extract, summarise or score, with {{field}} placeholders, and returns the new columns you define in outputFields for each row. Uses Claude Haiku. $0.002 per row. Runs nerolabs/dataset-ai-enrich.

Compares an old and a new list of JSON rows by a key field and returns what was added, removed and changed, with the old and new value of every changed field. Use it to spot new leads, price changes or deleted items between two scrapes. Up to 500 rows a side. $0.001 per new row sent (old rows free), $0.002 minimum. Same engine as nerolabs/dataset-diff-detector.

Works out every data breach notice a company owes and the deadline for each. Send how many people were affected per US state or country, the data types exposed, whether it was encrypted, the discovery date and flags (SEC registrant, HIPAA, GLBA, NYDFS). Returns each notice owed (people, attorneys general, credit bureaus, HHS, SEC 8-K, FTC, EU and UK regulators, Canada, Australia) with a dated dead

Checks a job post against pay transparency laws before it is published. Send the locations (and whether it is remote), employer headcount and posting date. Returns each law that applies or may apply, a merged checklist of what the post must show (pay range, benefits), salary history bans, penalties and the official source per law. Covers US states and cities, Canadian provinces and the EU Pay Tran

Returns the legal hourly minimum wage at a US address (or state plus city or county) on a date: the highest of the federal, state, county and city rates, which one sets it, the tipped cash wage, special industry rates, the next scheduled change and the government source per figure. Use it for payroll, job offers or wage checks. Unclear cases give the highest rate that could apply plus what it depe

Checks which US states a remote seller must register in to collect sales tax (economic nexus). Send sales and order counts per state for the last 12 months or a calendar year, and whether they include marketplace sales (Amazon, Etsy). Returns states over the threshold (must register), close to it (80 percent or more) and under it, with each state's threshold, measurement period, what counts and th

Extracts the text (plain or Markdown) and tables from up to 10 PDF links, with page count, title, author and dates, and flags PDFs that are scans with no text layer (use /ocr for those). Choose a page range and a length limit. $0.01 per PDF. Runs nerolabs/dataset-pdf-extract.

Checks up to 100 phone numbers from any country: valid or not, the number in international E.164 format (+441174960012) and national format, country, and line type (mobile, landline, toll free, VoIP) from Google's libphonenumber data. Set defaultRegion (GB, US...) for numbers written without a country code. Does not call the number. $0.002 per number, $0.002 minimum. Same engine as nerolabs/phone-

Looks up Irish companies in the official CRO register by company number or by name: status, type, registered address, incorporation date and recent filings. Up to 20 results a call. $0.01 per company (a name search costs maxSearchResults x $0.01, default 5). Runs nerolabs/ireland-cro-monitor.

Looks up Polish companies in the official KRS court register by KRS number: name, legal form, status, NIP and REGON, address, share capital, board members and registration dates. Up to 20 a call. $0.01 per company. Runs nerolabs/poland-krs-monitor.

Looks up UK companies in Companies House by company number: name, status, type, registered address, incorporation date, SIC codes, accounts and confirmation statement due dates, and officers. Up to 20 a call. $0.01 per company. Runs nerolabs/uk-companies-house-monitor.

Visits up to 10 business websites (home and contact pages) and returns the published contact emails (role inboxes such as info@ by default), phone numbers and social profiles for each. Honours robots.txt; nothing is guessed. $0.01 per website. Runs nerolabs/website-contact-finder; takes up to a minute.

Looks up up to 15 domains in the official RDAP registries (the modern WHOIS): when each was registered and how old it is, when it expires, the registrar, status codes and name servers. Flags new domains and ones about to expire. Useful for fraud checks, lead qualification and spotting expiring domains. $0.005 per domain. Same engine as nerolabs/domain-whois-checker.

Reads the text in images (PNG, JPEG, WebP, TIFF and more) and scanned or born-digital PDFs with OCR, in 100+ languages, and returns the text per file and per page with confidence scores. Up to 10 files a call. Priced per page allowance: files x maxPagesPerFile (default 5, at most 20) x $0.005. Runs nerolabs/bulk-ocr-image-pdf-to-text; can take a minute or two.

Runs Google PageSpeed Insights (Lighthouse) on up to 10 pages and returns performance, accessibility, best-practices and SEO scores, Core Web Vitals (LCP, CLS, INP, TBT) and the top fixes to make each page faster. Mobile, desktop or both. $0.005 per page per strategy. Runs nerolabs/bulk-pagespeed-checker; takes 20 to 60 seconds.

Detects the technology behind up to 20 websites: CMS, ecommerce platform, analytics, tag managers, frameworks, hosting and CDN, payment and chat tools, with confidence per finding. Filter for sites using a given technology to qualify leads. $0.005 per website. Runs nerolabs/tech-stack-detector.

Cleans up to 150 messy JSON rows: removes duplicates (exact, normalized or fuzzy), flattens nested JSON, trims and fixes emails, phone numbers and URLs, strips HTML, turns number and true/false text into real types, drops empty fields, keeps, removes or renames columns. Returns clean rows plus a report of what changed. $0.002 per row sent, $0.002 minimum. Same engine as nerolabs/dataset-cleaner-ex

Joins two lists of JSON rows on one or more key fields, like a VLOOKUP or SQL join: inner, left, right, full or anti. Keys match exactly, or normalized (case, spacing and number-as-text differences ignored). Returns joined rows plus match counts and unmatched rows. Up to 500 rows a side. $0.002 per left row sent (right rows free), $0.002 minimum. Same engine as nerolabs/dataset-join-merge.

Reads up to 50 job post texts with AI and returns structured fields for each: salary range normalised to yearly and hourly in a currency, seniority, remote or hybrid, visa sponsorship, employment type, years of experience, and required skills. $0.003 per job. Runs nerolabs/job-post-enricher.

Groups up to 500 JSON rows and works out totals per group: count, sum, average, min, max, median, distinct count, first and last, with date buckets (day, week, month, quarter, year), a pivot table, sorting, top N and a totals row. Returns one row per group plus a summary. $0.001 per row sent, $0.002 minimum. Same engine as nerolabs/dataset-aggregate-pivot.

Filters, transforms, sorts and dedupes up to 500 JSON rows in one call: conditions (equals, contains, greater than, regex, empty), field renames, trims, case changes, number and date parsing, computed fields, sort, distinct by field, offset and limit. Returns the kept rows plus counts of rows dropped and why. $0.001 per row sent, $0.002 minimum. Same engine as the Apify actor nerolabs/dataset-filt

Reads up to 20 RSS, Atom or JSON feeds, or plain website addresses (the site's own feed is found automatically), and returns the newest items: title, link, date, author, summary, tags, image and podcast audio link. Filter by date (publishedAfter: "24 hours") and keywords. $0.005 per feed sent, up to 25 items each. A call where no feed could be read is not charged. Runs nerolabs/rss-feed-reader.

On-page SEO audit of up to 20 websites or pages: a 0-100 SEO score, A to F grade and the issues to fix, worst first (missing title or meta description, H1, noindex, HTTPS, mobile viewport, thin content, image alt text, Open Graph, structured data, sitemap, robots.txt), plus optional broken-link check. $0.005 per website, plus $0.0005 per link allowed when checkLinks is on. A call where no website

Transcribes up to 5 audio or video links (MP3, WAV, M4A, MP4 and more; Google Drive and Dropbox links work) to text, optionally with speakers, timestamps and SRT subtitles. Priced per minute allowance: files x maxMinutesPerFile (default 5, at most 10) x $0.01 ($0.02 in detailed mode). Runs nerolabs/bulk-transcription; can take a minute or two.

Checks up to 10 UK businesses or traders and returns a trust verdict for each: Companies House status, age, accounts and filings, directors, Gazette insolvency notices, whether the website matches the company, and food hygiene rating where relevant, with the reasons. Give a name, website, company number or postcode. $0.05 per business. Runs nerolabs/uk-business-trust-check.

Builds a risk report on up to 5 UK companies by number or name: Companies House profile, overdue accounts, charges, director history and disqualifications, Gazette insolvency notices and a scored list of red flags with the evidence, optionally as Markdown. Use it before a deal, a supplier contract or credit. $0.50 per company. Runs nerolabs/uk-company-risk-report.

Transcribes up to 5 call recordings and scores each with AI: greeting, needs found, booking or sale made, objections handled, missed opportunities, sentiment and a coaching summary, plus your own questions. Priced per minute allowance: files x maxMinutesPerFile (default 5, at most 10) x $0.04. Runs nerolabs/call-score; can take a minute or two.

Scores creators from rows an Instagram, TikTok or YouTube scraper already returned (or your own rows): engagement rate by followers and by views, typical rate for that size, fake follower warning signs with reasons, a 0 to 100 score, grade A to F and a shortlist, check or skip verdict. No scraping. $0.003 per creator found in the rows (8 videos of one creator = 1), 100 creators at most. Not charge

Scores up to 20 product ideas or keywords against live demand: how many tools on the Apify Store already do it, monthly users of the top tools, the leader and its failure rate, tools that gained most users this month, x402 agent tools for the same job with paying wallets, and a verdict (UNPROVEN, SMALL, OPEN, HEATING UP, CONTESTED, CROWDED) with the reason. $0.02 per idea. A call where no idea cou

Downloads up to 20 images and converts them to WebP (default), JPEG, PNG or AVIF, optionally resized to a maximum width or height, EXIF stripped and photos turned the right way up. Returns a link to each converted file with its measured width, height and size before and after. $0.002 per image converted; a call where nothing converts is not charged. Runs nerolabs/bulk-image-converter.

Finds the logo file (from schema.org data or the header), the best square icon (apple-touch-icon, web app manifest or favicon, largest first) and the favicon of up to 50 websites, each downloaded and checked to be a real image, with measured width, height and format, plus site name and brand theme colour. A Clearbit logo API alternative. $0.002 per website where something is found; a call where no

Writes up to 1,000 JSON rows to a Google Sheet (append, replace, or update rows matched on keyColumn, adding new keys) or reads a tab back as JSON (mode import). No OAuth: share the sheet with nerolabs-sheets@nero-labs-sheets.iam.gserviceaccount.com as Editor. Keeps column order, adds new fields on the right, waits out Google rate limits. $0.005 per call plus $0.0002 per row sent, or $0.0001 per r

Checks if AI crawlers can read up to 20 websites: GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and Meta-ExternalAgent by default (four more optional). Reads robots.txt, fetches each site as each bot vs a browser to catch firewall blocks, and checks llms.txt. Returns allowed or blocked per bot and why. $0.005 per website. Unreachable sites are not charged. Runs nerolabs/ai-crawler-access-checker

Sends up to 5 prompts to several AI assistants, each with its own live web search via official APIs (openai, perplexity, claude, gemini, grok), and returns each answer, its cited sources and sites, and which assistants mention or cite each brand or website in trackBrands. Price per prompt: openai $0.02, perplexity $0.01, claude $0.03, gemini $0.05, grok $0.05 (default openai + perplexity = $0.03).

Asks up to 5 shopping prompts to OpenAI with its web search tool (ChatGPT-style answers via the official API, not scraped from chatgpt.com) and returns the recommended products in order with prices, shops, links and cited sources, plus whether each brand, shop or website in trackBrands appears and at what position. Set country for local shops. $0.05 per answer. Failed answers are not charged. Runs

Scores each business in rows a Google Maps reviews scraper already returned (or your own rows): reviews in the last 90 days vs the 90 before, momentum (growing, slowing, stalled), owner reply rate, unanswered 1 and 2 star reviews, rating trend, tags such as neglected, a lead score 0 to 100 and a one-line pitch. No scraping. $0.004 per business, 100 at most. Not charged if none can be scored.

Ranks the skills, tools, licences and certificates named in job rows a jobs scraper already returned (LinkedIn, Indeed, Glassdoor, JobSpy, or your sheet): share of jobs asking for each per job title or role family, rising or falling against an older batch or posting dates, median pay with and without it (pay premium), and a learn this next line. 175+ skills, company boilerplate skipped, no AI. $0.