MCP-сервер NeuralVerge

Какие серверы из официального реестра MCP принадлежат NeuralVerge, отвечают ли они и что умеют. Всё это проверила программа, а не переписал человек со страницы продукта.

прочитано 2026-10-20

← Назад к NeuralVerge

Чей это сервер

Имя в реестреai.neuralverge/neuralvergeподтверждён в реестре MCP

Имя в реестре лежит в пространстве neuralverge.ai. Реестр пускает в такое пространство только того, кто доказал владение доменом через DNS или HTTP.

Что он ответилпроверено 2026-10-20

Состояниеотвечаетпроверено 2026-10-20
Инструменты342
Сервер называет себяneuralverge
Версия сервера1.0.0
Версия протокола2025-06-18

2026-10-20 сервер ответил на initialize и tools/list без входа. Первый ответ за 647 мс.

Стоит ли агенту его подключатьПока не ясно

Делают ли инструменты работу продукта, ещё не проверено.
Проверок пока 1 из 7, нужных, чтобы судить о надёжности.

После подключения

Описания занимают около 67082 токенов контекста: включайте только нужные задаче инструменты.

Вывод сделан правилами по замерам выше. Вероятный первый вызов выбрала модель по списку инструментов: прочитайте описание, прежде чем на него полагаться.

Замер каждый деньежедневных проверок: 1, первая 2026-10-20

Оценка INITE MCPчерез неделюдля оценки нужно 7 ежедневных проверок, пока 1
Отвечалидёт замерпроверок пока 1, долю покажем после 7
Медиана первого ответа533 мс
Место в контексте агента≈ 67 082 токеновописания всех инструментов (342), агент держит их в контексте на каждом ходу
Описания инструментовB · 83 из 100
Типизированные ответы0 из 342 инструментовобъявлен outputSchema: ответ приходит с типами и читается сразу

Параметры без описания: в 2 из 342
Не сказано, только читает инструмент или может удалять: 342 из 342

Что агент может с ним сделать

не ясно · Build prospect lists from funding and hiring signals
может · Check a company's ownership and filings against official registries
не ясно · Track competitor launches, pricing and reviews

Каждый сценарий продукта мы задали Jev вместе со списком инструментов сервера: может ли агент сделать это этими инструментами?

На что посмотреть

В этих описаниях нет ни одного из пяти тревожных признаков.

Программа проверяет сервер раз в день: initialize и tools/list, без модели. Токены - это длина описаний из расчёта примерно четыре символа на токен. Оценка - среднее шести проверок имени, описания, схемы и аннотаций каждого инструмента. Она говорит, насколько понятно описаны инструменты, а не о том, работают ли они. Оценка складывается только из замеренного: отвечает ли сервер (30%), делает ли работу продукта (25%), описания (20%), контекст (10%) и использование (15%). Незамеренная часть не учитывается, а скрытая инструкция ограничивает оценку 40 баллами. Состояние MCP-серверов вендоров: методика и цифры →

Инструменты сервераинструментов: 342 · прочитано 2026-10-20

run_search
Runs a synchronous web search and returns a structured list of ranked results (title, url, snippet) immediately, without polling.
run_extract
Loads a webpage by URL and extracts structured data from it according to natural-language instructions and/or a JSON schema. Synchronous — returns the result immediately.
run_linkedin_email
Looks up a LinkedIn profile by its full profile URL and returns contact details, including an email address when available.
run_linkedin_domain
Finds a LinkedIn profile matching a company name or domain plus a full name.
run_linkedin_company_search
Searches LinkedIn companies by a text query with optional filters (size, industry, location).
run_linkedin_people_search
Searches LinkedIn people using an optional free-text query and a rich set of advanced filters (company, title, seniority, location, experience, industry).
run_linkedin_company_employee
Searches LinkedIn employees of one or more given companies, using the same people-style filters as run_linkedin_people_search.
run_email_enrichment
Enriches a known email address with associated profile data (name, phone, company, position, LinkedIn/X profiles) when available.
run_email_validation
Validates an email address and returns deliverability signals (valid/invalid/risky, catch-all detection, provider, confidence).
run_email_finder
Finds a professional email address for a person given their company domain, first name, and last name.
run_phone_enrichment
Enriches a known phone number with associated profile data (name, email, company, LinkedIn/X profiles, carrier) when available.
run_phone_enrichment_us
Validates and enriches a US phone number with carrier, line type, activity score, litigator risk, and owner details.
run_crunchbase_company
Fetches structured company data (website, location, founding year, employees, industries, funding, description) from a Crunchbase company URL.
run_alibaba_product
Get full Alibaba product details by ID or URL: price tiers, MOQ, attributes, variants, and supplier. Flat 5 points per successful run.
run_alibaba_product_search
Search Alibaba products by keyword: price, MOQ, supplier, and certifications. 1 point per returned record.
run_aliexpress_product
Get AliExpress product details by ID: title, price, images, and store. Flat 5 points per successful run.
run_aliexpress_product_reviews
Get buyer reviews for an AliExpress product by ID: rating, text, and buyer details. 1 point per returned record.
run_aliexpress_product_search
Search AliExpress products by keyword: title, price, and image. 1 point per returned record.
run_amazon_product
Get full Amazon product details by ASIN: title, brand, price, images, rating, features, and seller. Flat 5 points per successful run.
run_amazon_product_offers
Get all marketplace offers for an Amazon product by ASIN: price, condition, seller, and delivery. Flat 5 points per successful run.
run_amazon_product_search
Search Amazon and extract product listings (title, URL, price, rating) across paginated results. 1 point per returned record.
run_amazon_seller
Get an Amazon seller profile by seller ID: name, rating, business details, and contact info. Flat 5 points per successful run.
run_amazon_seller_products
List products sold by an Amazon seller's storefront: title, URL, price, rating, and image. 1 point per returned record.
run_ar_companies
Get one company from the Argentina registry of legal entities (Registro Nacional de Sociedades, IGJ) by CUIT or registry link: name, company type, contract date, legal and fiscal address, activities and last update. 5 point per returned…
run_ar_companies_search
Search the Argentina registry of legal entities (Registro Nacional de Sociedades, IGJ) by company name or CUIT: name, CUIT, company type, address, main activity and link. 1 point per returned record.
run_at_companies
Get one company from the Austria registry (Firmenbuch) by registration number (FN) or registry URL: name, status, legal form, address, previous names and the paid extracts offered. Flat 5 points per successful run.
run_at_companies_search
Search the Austria company registry (Firmenbuch) by company name or FN: name, registration number, status, seat and link. 1 point per returned record.
run_au_companies
Get one company from the Australia registry (ASIC / ABN) by registration number or registry URL: name, status, registration date, address and legal form. Flat 10 points per successful run.
run_au_companies_search
Search the Australia company registry (ASIC / ABN) by company name: name, registration number, status, address and link. 1 point per returned record.
run_avito_categories
Get the Avito category tree: departments and sub-categories with id, name, alias (URL slug for the category filter), parent id, level and browse URL. 5 point per returned record.
run_avito_items
Get a full Avito listing by id or URL: title, price (RUB), description, specification params, category and breadcrumbs, address and coordinates, view counts, posting time, photos and seller. Flat 5 points per successful run.
run_avito_items_search
Search Avito listings by keyword with region, category, price range, sort, seller type, Avito Delivery and attribute filters. Returns title, price, location, coordinates, image and seller for each listing. 1 point per returned record.
run_avito_sellers
Get an Avito seller (brand page) by slug or URL: name, type, rating, review count, followers, active listing count, registration date and badges. Flat 5 points per successful run.
run_be_companies
Get one company from the Belgium registry (KBO/BCE) by enterprise number or registry URL: name, status, start date, address, legal form, activities, directors, contact details and establishment units. 10 point per returned record.
run_be_companies_search
Search the Belgium company registry (KBO/BCE) by company name: name, enterprise number, status, address and link. 1 point per returned record.
run_betalist_startups
Get one startup from BetaList by slug or URL: name, tagline, description, makers, topics, launch date, screenshots and similar startups. Flat 5 points per successful run.
run_betalist_startups_latest
The newest startups on BetaList, newest first: name, tagline, logo and link. 1 point per returned record.
run_betalist_startups_search
Search BetaList startups by keyword: name, tagline, logo and link. 1 point per returned record.
run_betalist_topics_startups
Startups of one BetaList topic (slug like generative-ai, category/topic, or a BetaList topic URL): name, tagline, logo and link. 1 point per returned record.
run_bg_companies
Get one company from the Bulgaria registry (Commercial Register) by UIC (ЕИК) or registry link: name, status, first registration date, address and contacts, legal form, activity, capital and shares, representatives and board, procuracy…
run_bg_companies_search
Search the Bulgaria company registry (Commercial Register) by company name or UIC: name, UIC (ЕИК), legal form and card link. 1 point per returned record.
run_bg_company_filings
List the announced acts of a Bulgaria company (Commercial Register) by UIC (ЕИК) or registry URL, newest first: announcement date, act type, title and document link (annual reports, notices, reports, current articles). 1 point per…
run_br_companies
Get one company from the Brazil registry (CNPJ / Receita Federal) by registration number or registry URL: name, status, registration date, address and legal form. Flat 5 points per successful run.
run_br_companies_search
Search the Brazil company registry (CNPJ / Receita Federal data) by company name: name, CNPJ, status and link. 1 point per returned record.
run_ca_companies
Get one federal corporation from the Corporations Canada registry by corporation number or registry link: name, status, incorporation date, registered office address, governing legislation, directors, individuals with significant control…
run_ca_companies_search
Search federal corporations in the Corporations Canada registry by name, corporation number or business number: name, corporation number, status, business number and link. 1 point per returned record.
run_ca_company_filings
Certificates and filings of one federal corporation in the Corporations Canada registry: date, type and amendment details. 1 point per returned record.
run_capterra_categories
List Capterra software categories: name, alias, and canonical URL. Flat 5 points per successful run.
run_capterra_category_products
List Capterra products within a software category, ranked as Capterra ranks them. 1 point per returned record.
run_capterra_product
Get full Capterra product details by id or URL: vendor, pricing, ratings, and features. Flat 5 points per successful run.
run_capterra_products_reviews
Get user reviews for a Capterra product by id or URL: rating, pros/cons, and author. 1 point per returned record.
run_capterra_products_search
Search Capterra software products by keyword: name, rating, review count, and image. 1 point per returned record.
run_co_companies
Get one company from the Colombia registry (RUES, Registro Mercantil) by NIT or registry URL: name, status, registration date, address, legal form, economic activity, legal representation and establishments. Flat 5 points per successful…
run_co_companies_search
Search the Colombia company registry (RUES, Registro Mercantil) by name: name, NIT, status, chamber of commerce, category and link. 1 point per returned record.
run_craigslist_areas
Directory of ~700 Craigslist regional sites with alias (the area value for searches), area code, hostname, country, region, coordinates, timezone and sub-areas. Flat 5 points per successful run.
run_craigslist_categories
Every Craigslist posting category with its alias (the category value for searches), description and section. Flat 5 points per successful run.
run_craigslist_postings
Full Craigslist posting by id or URL: title, price, body text/HTML, category attributes with matching search filters, decoded VIN spec for vehicles, hiring company for jobs, images, location and dates. Flat 5 points per successful run.
run_craigslist_postings_search
Search live Craigslist postings in any regional site by keyword, category, area/sub-area/postal/coordinates and 60+ vehicle, motorcycle, boat, housing, job, gig and event filters; returns title, price, date, location, images and category…
run_craigslist_postings_seller
All postings published by the seller of a given Craigslist posting (a dealer's, landlord's or shop's full inventory). 1 point per returned record.
run_cy_companies
Get one organisation from the Cyprus registry (Registrar of Companies, DRCOR) by registration number (HE, EE, AE, S, BN prefix) or registry link: name, status, registration date, registered office, type, previous names, directors /…

Первые 60 из 342, в порядке сервера.

Как подключить

Добавьте сервер в конфиг MCP-клиента. Claude Desktop, Cursor и большинство других клиентов понимают такой вид:

{
  "mcpServers": {
    "neuralverge": {
      "type": "http",
      "url": "https://api.neuralverge.ai/functions/v1/mcp-server"
    }
  }
}

Подключить

Claude Code, в терминале

claude mcp add --transport http neuralverge https://api.neuralverge.ai/functions/v1/mcp-server

Добавить в Cursor Добавить в VS Code

claude.ai, ChatGPT и другие клиенты: добавьте https://api.neuralverge.ai/functions/v1/mcp-server как свой коннектор в настройках.

Каждую строку нашла программа: запись в реестре, ответ самого сервера, спецификация, которая разбирается. Если строки нет, значит, мы этого не нашли, а не что этого нет.