Company Facts

dev.companyfacts由 companyfacts.dev 發布官方發布者v1.1.0更新於 Nov 7, 2026

Company data from official sources, plus a finder for the best paid x402 API for any task.

已驗證Streamable HTTP可網頁執行網頁搜尋與抓取金融商業與電商

概覽

AI 產生的概覽

遠端 MCP 伺服器,從官方來源提供公司資料:職缺看板、企業登記、DNS、SEC 財報,以及通用網頁讀取工具。

可以做什麼
  • 查找並讀取職缺:找出公司使用的招募系統、列出開放職缺、跨最多 20 家公司比較、逐日追蹤職缺變化並彙整招募狀況。
  • 查詢公司:州登記機關的新設企業紀錄、依國家檢索企業登記、VAT 號碼驗證、SEC 財務資料、聯邦合約。
  • 檢查網域與網站:RDAP 註冊資訊、DNS、郵件服務商、SPF/DMARC、技術堆疊辨識。
  • 讀取網頁:將網頁與 PDF 轉為 markdown、依指定欄位擷取為 JSON、檢查電子郵件地址。
  • 檢索外部目錄:學術論文、Hacker News、App Store 應用程式,以及付費 x402 API 端點。
適用情境
適合需要可查證的公司事實(招募動態、登記資料、財務數字、網域與技術細節),或一般網頁讀取與欄位擷取的場合,不必自行撰寫爬蟲或串接各廠商 API。
執行需求
遠端 Streamable HTTP 端點,無需安裝。每天 10 次免費呼叫,不需帳號或金鑰;超過後可透過 x402(Base 上的 USDC)按次付費,或使用預付金鑰。extract_page 需要 API 金鑰。另有僅含三個職缺工具的本地 PyPI 版本。
安裝前請注意
超過每日免費額度後需付費,可按次以 USDC 支付或使用預付金鑰;extract_page 會消耗付費餘額。工具標示為唯讀,但 read_page 與 extract_page 會抓取第三方網頁,所請求的網址與擷取內容會離開本機。依設計不傳回個人資料。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 Company Facts,將其新增到工作區。
  2. 為需要使用其工具的對話啟用該服務。

Web executable,透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。

交給代理程式安裝

把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式,或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南,先向你展示發布者、端點和登入方式,經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。

閱讀 https://sourceweft.com/mcp/dev.companyfacts%2Fcompany-data/install.md,並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前,先告訴我它的發布者、端點和登入方式,等我確認。不要把 API 金鑰寫進命令或我會提交的檔案;問我金鑰應從哪裡取得。

檢視代理程式所依據的安裝指南

其他 MCP 客戶端

把它新增到你客戶端的 mcpServers 設定中。

{
  "mcpServers": {
    "company-facts": {
      "type": "http",
      "url": "https://companyfacts.dev/mcp/"
    }
  }
}

README

ats-jobs-mcp: Company Facts

An MCP server for company data read from the official or original source. Newly formed US businesses straight from state registers, with their website found only on proof. A company's register entry, tech stack, DNS and email provider, SEC financials and open jobs, each with a link to where it came from.

The full server runs remotely at https://companyfacts.dev/mcp with 20 tools and nothing to install. This PyPI package is the small local edition: the three job tools, run on your own machine.


Use it with no install

Point any MCP client that speaks Streamable HTTP at the URL:

json
{  "mcpServers": {    "company-facts": {      "url": "https://companyfacts.dev/mcp"    }  }}

No account and no key: 10 free calls a day. Past that, pay per call with x402 (USDC on Base) or use a prepaid key (see https://companyfacts.dev/pricing). A call that finds nothing is not charged.

The same tools answer plain HTTP, for agents that do not speak MCP:

  • GET https://companyfacts.dev/openapi.json for a tool-calling agent
  • GET https://companyfacts.dev/llms.txt for a model reading the service in prose
  • GET https://companyfacts.dev/v1/new-businesses?states=CO&mode=new

Company tools (remote)

ToolWhat it returns
new_businessesNewly formed businesses from the official Secretary of State registers of New York, Colorado, Oregon, Connecticut and Pennsylvania, in one record format. Feed by formation date, search by name, or look up by state entity number (KYB checks). Each record carries the company's website when one was found with proof, else says none was found.
company_searchSearch every company we track: filter by state, city, name, formation date, technology on its site (Shopify, WordPress, ...), whether it publishes a phone, role inbox or contact page, and hiring. A search with no match is not charged.
company_lookupOne company from its domain: registrar, DNS and email provider, website tech stack, hiring from its own job board, plus its official register entry or SEC figures when you give name and country or a ticker.
search_company_registryOne country's official company register by name (UK, Norway, France, Finland, Czechia, Slovakia, Ireland, Australia, or GLEIF worldwide), in one shape.
sec_financialsUS public company financials from SEC EDGAR XBRL.
validate_vatAn EU VAT number checked against the European Commission's VIES service.
domain_lookupRegistration (RDAP), DNS records, email provider, SPF and DMARC for a domain.
detect_tech_stackThe technologies behind a web page, each with its evidence.
federal_contractsAwarded US federal contracts from USAspending.

Websites are attached to new businesses only on proof: the site's own pages must show the exact legal name and the same city, state or ZIP as the register. Records never carry officers, directors, individual registered agents, emails, phones or street addresses.

Job tools

ToolWhat it returns
find_company_boardWhich applicant-tracking system a company's careers page runs on, and its board name.
list_open_jobsEvery open job at one company, read from its own Greenhouse, Ashby, Lever or Workable board.
compare_companiesOpen jobs across up to 20 companies in one schema.
job_changesJobs newly posted, taken down, reopened or edited, day by day, from a daily record of every board.
hiring_summaryOne company's hiring at a glance: roles by department and place, remote share, salary ranges, recent openings and closings.

The vendors publish open job board APIs, and all of them are useless unless you already know the company's board name. None publishes a directory of its customers, so this carries one: 8,532 boards on the remote server, each confirmed against the vendor's own API.

Other tools (remote)

read_page (a web page or PDF as markdown), extract_page (named fields from a page as JSON, paid only), check_email (syntax, mail servers, disposable and role flags), search_papers (OpenAlex), search_hacker_news, app_store_lookup.

Every tool is read-only and declares a title and readOnlyHint.

Salaries, parsed, and refused when unclear

Ashby prints pay as "$211.4K - $290.6K". Greenhouse prints it in cents, sometimes as several bands at once. Both are turned into numbers:

json
"salary_min": 211400.0, "salary_max": 290600.0,"salary_currency": "USD", "salary_text": "$211.4K - $290.6K"

The parser refuses more often than it guesses, deliberately:

  • $50 - $80 / hour gives no salary. An hourly rate is a different quantity and reporting it as annual would be a confident lie.
  • 600 000 kr gives no salary. kr is Swedish, Norwegian and Danish.
  • 0.5% - 1.75% - Offers Equity gives no salary. Equity is not pay.
  • Bands in two currencies are never spanned into a range that exists nowhere.

salary_text always keeps what the employer actually wrote.

No personal data

Recruiter names are never returned. Email addresses and phone numbers are removed from job descriptions. Unlisted Ashby postings are dropped, because Ashby's documentation says they should not be shown publicly.

Or install the local job edition

bash
pip install ats-jobs-mcp
json
{  "mcpServers": {    "ats-jobs": {      "command": "ats-jobs-mcp"    }  }}

The local edition has find_company_board, list_open_jobs and compare_companies over a bundled directory of 7,479 Greenhouse and Ashby boards (Lever by name, for example lever:spotify). It is free, runs on your machine, and takes no cut of anything.

Python and LangChain

  • pip install groundtruth-company: a Python client for the company tools.
  • pip install langchain-groundtruth: the same tools as LangChain tools.

Relation to the Apify Actors

The same job parsing runs as an Apify Actor at apify.com/groundtruth/ats-job-listings. tests/test_core_parity.py fails if the two copies of the parsing logic drift apart.

Licence

MIT. Job postings belong to the employers who published them. Register records belong to the states that publish them.

來源:README.md,提交 1b840c4

工具

21

find_company_board

Find a company's job board

Find which applicant-tracking system a company's careers page runs on. Greenhouse, Ashby, Lever and Workable all publish open job board APIs, and all of them are unusable unless you already know the company's board name. None publishes a directory of its customers, and none could, since no vendor knows about the others. Use this first when you have a company name rather than a careers URL.

輸入參數limitcompany必填

list_open_jobs

List a company's open jobs

Every open job at one company, read from its own careers system. Returns title, department, location, employment type, workplace type, posted date, a parsed salary range where the employer publishes one, and a verify_url pointing at the employer's own advert. Salaries are refused rather than guessed: an hourly rate, an ambiguous currency, or bands in two currencies all yield no figure, and salary_text keeps whatever the employer wrote.

輸入參數limitcompany必填include_description

compare_companies

Compare several companies' jobs

Open jobs across several companies at once, in one schema. This is the thing no single applicant-tracking system can do: each vendor only knows about its own customers. Companies on different systems come back with the same field names. One company failing does not lose the others.

輸入參數companies必填per_company

job_changes

Jobs opened, closed or changed

What changed in hiring, day by day: jobs newly posted, taken down, reopened, or edited (title, location or salary), read daily from every job board in the directory. Call it each day with since set to the last day you saw, rather than re-reading every board. Omit companies for the whole market, or filter by kind and title words.

輸入參數qkindslimitsinceuntiloffsetcompanies

read_page

Read a web page

Any public web page or PDF as clean markdown, with its title and final URL after redirects, plus the page's schema.org data (prices, dates) when it has any. robots.txt is honored. Some sites refuse automated readers; those calls fail with the reason and cost nothing.

輸入參數url必填max_chars

extract_page

Pull structured data from a page

Read a web page and return the named fields as JSON, using only what the page states; a value the page does not give comes back empty rather than guessed. Name fields, or pass a JSON Schema, or neither for a title, summary and key points. Needs an API key (5 calls from its balance); a failed extraction is refunded.

輸入參數url必填fieldsschemainstructions

check_email

Check an email address

Is this address worth sending to: syntax, whether the domain exists and has mail servers, and whether it is a throwaway, role (info@, sales@) or free-provider address. It never contacts the mailbox, so a missing mailbox at a real domain still passes, and the answer says so.

輸入參數email必填

domain_lookup

Look up a domain (registration, DNS, email provider)

For one domain: whether it is registered, its registrar, created, updated and expiry dates, status codes and nameservers (from the registry's RDAP server), plus its A, AAAA, MX, NS and TXT records, the email provider behind its MX, and whether it publishes SPF and DMARC. No registrant names or contact details are returned.

輸入參數domain必填

detect_tech_stack

Detect a website's tech stack

Fetch one web page and report the technologies behind it (CMS, ecommerce platform, analytics, CDN, web server, frameworks, payment processors, email provider) with a confidence and the evidence for each. Reads HTML, headers, cookies and DNS from one page load; it cannot see what only a running browser reveals.

輸入參數url必填min_confidence

validate_vat

Validate an EU VAT number

Checks one VAT number against the European Commission's VIES service and returns whether it is a live registration and, where the member state shares it, the registered name. Covers the 27 EU states and Northern Ireland (XI). A busy or down member state is reported as an error, never as an invalid number.

輸入參數vat必填

search_company_registry

Search an official company registry

Searches one country's official company register by name and returns registered companies in one consistent shape: number, legal name, status, legal form, dates, town and postcode, and a link to the register entry. Covers GB, NO, FR, FI, CZ, SK, IE, AU, the worldwide LEI register, and US entities in any state that hold an LEI (with their state filing number). Companies only, no directors or contact details.

輸入參數name必填limitstatecountry必填

search_papers

Search academic papers

Search the OpenAlex catalogue of scholarly papers by keyword, look papers up by DOI, list the papers that cite one work, or list a university's or journal's papers. Returns title, authors, year, citation count, DOI, abstract and a free full-text link when one exists.

輸入參數doismodesortquerysourceto_yearfrom_yearcited_workwork_typesinstitutionmax_resultssearch_scope另有 4 個

search_hacker_news

Search Hacker News

Search Hacker News stories and comments with date, points and comment-count filters, read the live front page and the top, best, new, ask and show lists in rank order, or read whole comment threads. Every item links back to news.ycombinator.com.

輸入參數modequerydate_tosort_bydate_frommax_itemssearch_instory_idsstory_tagmin_pointscontent_typemin_comments另有 1 個

sec_financials

SEC financials and screener

Reads US public company financials from SEC EDGAR XBRL. Company mode returns one company's reported history of a metric (revenue, net income, assets and 20 more, or any us-gaap tag) by ticker or CIK. Screen mode ranks every filer on one metric for one calendar period. Every figure carries period_type and a link to its filing. Reports figures, not advice.

輸入參數modelimitordermetricperiodcompanymax_valuemin_valueperiod_type

hiring_summary

Summarise a company's hiring

One company's hiring at a glance: open roles counted by department and by location, remote share, salary ranges per currency where the job board publishes pay, and which roles opened or closed in the last 7 or 30 days from a daily record of every board. Reads the company's live Greenhouse, Ashby, Lever or Workable board. Give a company name, board name or careers URL.

輸入參數dayscompany必填

federal_contracts

Federal contract awards and recompetes

Searches awarded US federal contracts on USAspending.gov by agency, keyword and dollar amount, largest first. Recompete mode instead lists contracts expiring soonest within a window (set min_amount to keep the search narrow), with the incumbent and days until expiry. Every record links to the official award page.

輸入參數modelimitagencykeywordend_datemin_amountstart_dateaward_typesexpiring_after_daysexpiring_within_days

app_store_lookup

Look up Apple App Store apps

Search the Apple App Store, look apps up by id or bundle id, read customer reviews, or get the top free, paid and grossing charts, per country. Returns name, developer, price, rating, rating count, version, genres and the App Store link.

輸入參數idsmodepagetermchartcountrycategorybundle_idmax_resultsreview_sortinclude_description

company_lookup

Everything public about a company

One company in one call, from its domain: registrar, DNS and email provider, the website's tech stack, and hiring from its own job board (open roles by department and place, salary ranges, roles opened or closed lately). Add name and country for the official register entry, or ticker for SEC figures. Each part links to its source; a part with nothing to report says why.

輸入參數namedomain必填tickercountry

new_businesses

Newly formed US businesses and state register lookup

Searches the official Secretary of State business registers of New York, Colorado, Oregon, Connecticut and Pennsylvania, in one record format. mode new is a feed of entities formed in a date window, newest first: use it for new-business leads (LLCs and corporations that just registered) and prospecting. mode search finds entities by legal name and mode lookup by the state's own entity number: use them for KYB (know your business) checks on a supplier or counterparty, to confirm it exists, its type, status (active, delinquent, inactive) and formation date. Every record links to the state's own data row. Never returns officers, directors, individual registered agents, emails, phones or street addresses: business city, state and ZIP only. Each record carries the company's website when one was found with proof (its own pages show the exact legal name and the same city or state), else 'no website found'. New York, Oregon and Pennsylvania publish active entities only, so a missing entity there may be inactive or unregistered. Pennsylvania publishes monthly, so its newest entities are weeks old. If one state fails the others still return and the failure is listed in missing.

輸入參數modenamelimitsinceuntilstatesstatusentity_typeentity_numbers

find_x402_api

Find the best x402 API for a task

Say what you need in plain words ('verify an email', 'read a web page', 'company record for a domain') and get the x402 endpoints that do it, best first, ranked only on public facts: wallets that paid it more than once in the last 30 days (Coinbase's CDP catalog), whether it answered our unpaid request with a 402 last night, its price, and PayAI's uptime. Each result has the price, networks, method, and links to re-check every number. No match is never charged.

輸入參數need必填sortlimitnetworkinclude_downmax_price_usd

版本歷史

1
  1. v1.1.0最新Oct 31, 2026

評分

已滿足適用檢查項的 100%

該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通必需25 分
  • 已宣告認證方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分

其他 Companyfacts 整合