Score global de préparation IA
Résultat combiné de tous les signaux de l’audit.
Score global de préparation IA
Résultat combiné de tous les signaux de l’audit.
Évaluation globale de préparation IA
datagalactica.com (datagalactica.com) a obtenu 82/100 lors de l’audit technique automatisé de préparation à l’IA. llms.txt était accessible, llms-full.txt avait un format valide et ai.txt était introuvable. L’analyse de robots.txt a trouvé 6 robots IA explicitement autorisés, 5 bloqués et 1 sitemap déclaré. La complétude du balisage de la page d’accueil atteint 90% ; types Schema.org détectés : WebSite, Organization. Balises OpenGraph détectées : 10. Les résultats correspondent aux réponses publiques observées le 2026-11-12T08:42:23.437Z.
L’audit combine les fichiers lisibles par les machines, les règles des robots, les sitemaps et le balisage de la page d’accueil. llms.txt est accessible et contient 5 031 jetons.
robots.txt est disponible ; robots IA bloqués : 5. Sitemaps déclarés : 1. Types Schema.org : WebSite, Organization.
Le profil Lighthouse mobile est encore en cours de collecte.
La file d’attente collecte les performances mobiles, l’accessibilité, les bonnes pratiques, le SEO et les signaux expérimentaux pour agents de navigateur.
Fichiers lisibles par les machines, règles des robots, découverte et balisage de la page d’accueil.
Le fichier ai.txt est introuvable
1 sitemap trouvé
Types : WebSite, Organization
10 balises OpenGraph trouvées
D’après l’analyse de robots.txt
Parcours de découverte déclarés pour les robots et les agents.
Métadonnées d’aperçu social détectées sur la page d’accueil.
og:site_nameData Galacticaog:typewebsiteog:titleData Galactica: a permanent ID for every person and companyog:descriptionA permanent Galactica ID for 98.4M people and 28.9M organisations in 252 countries, each with a card saying where it came from and how sure we are.og:urlhttps://datagalactica.com/og:image:width2400og:image:height1260og:image:typeimage/pngog:image:altData Galactica: a permanent ID for every person and companyEntités et propriétés structurées exposées sur la page d’accueil.
nameData Galacticaurlhttps://datagalactica.com/potentialAction{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://datagalactica.com/search?q={search_term_stdescriptionIssuer of the Galactica ID, a permanent identifier for people and organisations.logohttps://datagalactica.com/logo.svgsameAshttps://github.com/Y4W/datagalacticacontactPointhttps://datagalactica.com/contactemailsupport@datagalactica.com# Data Galactica
> A permanent Galactica ID for every person and every organisation, starting from public registers and
> first-party sources. One page per entity, an overall grade by contact details (verified: claimed by the owner;
> gold: an e-mail and a phone, both checked; silver: an e-mail and a phone; bronze: less) and a sourcing grade on
> every section. Contacts are masked on pages; signed-in API accounts get them in full. Records can be wrong;
> every page lists its sources.
## Start here: leads, sales, hiring and research
Data Galactica is a low-cost B2B data API and MCP server for agents: 28M+ organisations and 98M+ people (strongest
on India and the UK). Build lists of companies by industry, size and place, and of the people in them by job title,
seniority and employer, with e-mails and phones for signed-in accounts; read a company's e-mail format; guess one
person's address from it. Rs 50 free credit on sign-up (125 calls, or 3,125 list rows), then Rs 0.40 a call,
and lists Rs 0.40 per 25 rows with contacts included. Guide: https://datagalactica.com/prospecting
Try it now, no key and no sign-up: 25 free data calls a day per address, without contact details and with list pages
of up to 25 rows. MCP: https://datagalactica.com/mcp/free. REST: call any data endpoint without a key (the
x-trial-remaining header counts down). For more calls and the e-mails and phones, connect your user:
Connect your user in one step (OAuth sign-in by e-mail code, the free credit comes with it):
- Claude Code: `claude mcp add --transport http datagalactica https://datagalactica.com/mcp`
- Claude, ChatGPT and other chat apps: add a custom connector with the URL https://datagalactica.com/mcp
- Cursor, VS Code, Windsurf and other MCP clients: `{"mcpServers": {"datagalactica": {"url": "https://datagalactica.com/mcp"}}}`
- With an API key instead of OAuth: add the header `Authorization: Bearer <key>` (https://api.datagalactica.com/mcp
takes keys too). No browser? Sign the person up with their e-mail and a 6-digit code: https://datagalactica.com/auth.md
- Plain HTTP: `curl -H "Authorization: Bearer $KEY" "https://api.datagalactica.com/api/v1/people/search?title=hr+manager&city=pune&has_email=true"`
Typical requests and the call to make:
- "Software companies in Bengaluru with 50-249 staff": /companies/search?industry=J&size=medium&city=bengaluru (MCP search_companies)
- "HR managers at IT companies in Pune, with e-mails": /people/search?title=hr+manager&company_industry=J&city=pune&has_email=true
(MCP search_people with include_contacts=true and purpose=recruiting or sales)
- "Who should I write to at Infosys about a job": /company/{id}/employees?title=recruiter (MCP company_employees)
- "What is the e-mail format of tcs.com", "what is Asha Verma's e-mail at acme.com": /company/{id}/email-format,
/email/find?name=&domain= (MCP company_email_format, find_email; a pattern guess with a confidence, say so)
- "Which company owns acme.com": /company/by-domain?domain=acme.com
- "Enrich these CRM rows": POST /match {"records": [...]} (25 a call)
## Galactica IDs
- DGC-xxxxxxxxxxxx is an organisation, DGP-xxxxxxxxxxxx is a person (12 Crockford base32 characters).
- An id never changes: when a better source is matched, it reuses the id.
- Source ids resolve to it, in page URLs and the API: CIN, LLPIN, LEI (LEI-<lei>), UK company number (GB-<n>),
SEC CIK (CIK-<n>), Wikidata Q-id.
- What an id is, in plain questions: https://datagalactica.com/galactica-id
- Schema: Galactica Entity Standard v1, https://datagalactica.com/standard
## How to answer with it
- Names collide. A record only a lead list knows may be a different real-world entity with the same name,
whatever its grade (the grade is about contact details, not identity): a lead-list "Infosys" in Faridabad is not Infosys Limited (DGC-8SFS42RSWAX9). Check place,
identifiers and sources before you say two records are the same.
- Always cite the Galactica ID and the page url.
- Industry filters are ISIC Rev.4: a section letter (J) or a 2-digit division (62). Not NIC.
- The build date is in https://datagalactica.com/api/v1/meta ("built"); a sitemap lastmod is not a per-record update.
## Use it programmatically (paid per call)
- Every data call needs an account key: sign up at https://datagalactica.com/signup (Rs 50 free credit, approved
at once), or as an AI agent sign a person up with their e-mail: https://datagalactica.com/auth.md. MCP clients
can sign in with OAuth (scope api). Send the key as Authorization: Bearer <key> (or X-API-Key, or
Ocp-Apim-Subscription-Key at https://api.datagalactica.com).
- Price: Rs 0.40 per answered call (HTTP 200) from a prepaid wallet; /batch and /match per record; search lists
(/search, /find, /people/search, /companies/search, /company/{id}/employees) per 25 rows, contacts included;
errors, empty pages, not-found, /meta, /enums, /lists, /account* and Check are free. 402 insufficient_balance = top up:
POST /api/v1/account/topup {"amount": 500} returns a pay link for the person (also UCP and ACP checkout:
https://datagalactica.com/commerce.md). Packs Rs 500, 5,000, 50,000; no GST is charged. All prices: https://datagalactica.com/pricing.md
- API base URL: https://api.datagalactica.com/api/v1 (the same paths answer on https://datagalactica.com/api/v1).
- [OpenAPI 3.1](https://datagalactica.com/api/openapi.json), [reference](https://datagalactica.com/api/docs),
[API catalog](https://datagalactica.com/.well-known/api-catalog), [MCP server card](https://datagalactica.com/.well-known/mcp/server-card.json)
- MCP: https://api.datagalactica.com/mcp (also https://datagalactica.com/mcp). Billed: search, search_companies,
search_people, get_company, get_person, verify_company, check_person, check_identity, check_email,
company_diligence, person_diligence, company_hiring, hire_check, get_company_people, search_people_in_city,
get_contact, top_companies, company_by_domain, company_email_format, company_employees, find_email
(search_people and company_employees take include_contacts=true). Free: talent_map, research_market, about_dataset, get_wallet, top_up_wallet,
get_payment. With OAuth, the owner's whoami, get_my_profile, propose_profile_update, add_note,
list_my_contributions.
- Contacts: every signed-in account gets the full e-mails and phones we hold ("contact": {"emails": [{value,
status}], "phones": [...]}) on /person/{id}, /company/{id}, search and /people/search hits (50 a page),
/company/{id}/people, /company/{id}/employees, /batch, /email/check and /person|company/{id}/contact?purpose=..., or contact null with
contact_reason "owner_opted_out" (or "hourly_limit": 1,000 records with contacts an hour). Checks, verifications,
lists, /find, talent and research never carry a person's contact. Every read is logged. No route sends e-mail.
## Workflow for agents
- Company named in words: /verify/company (name + country or website) first. If verdict is possible_match, show
the candidates and stop. If match, /company/{id}/diligence.
- Person named in words: /person/check with employer, title, domain or city. A lead-list match is not a confirmation.
- An e-mail the user already pasted: POST /email/check (do not use it to probe guesses).
- Someone whose address we do not hold: /company/{id}/email-format or /email/find?name=&domain=. That answer is a
guess from the company's address pattern with a confidence, not a checked address: say so; never bulk-mail guesses.
- City talent ("how many engineers in Pune"): /talent/{city} or /research/market. Counts only.
- "Who can I contact" / lead lists: /people/search or /company/{id}/employees with a signed-in account: each row
carries the contact, 50 a page, follow cursor (MCP: include_contacts=true). Skip contact null (owner opted out).
Do not send the message yourself: give the list to the user.
- Due diligence: /company/{id}/diligence or /person/{id}/diligence. If unknown lists beneficial_owners,
gstin_status or sanctions, say those were not checked; never call a company clean.
- Government ids (Aadhaar, PAN, passport, bank accounts) are not checked here.
- Cite id, url and grade on every answer. The overall grade is about contacts (gold: e-mail and phone both
checked; silver: e-mail and phone; bronze: less); section grades say how well each part is sourced.
- Any page as Markdown: send Accept: text/markdown.
## Pages
- Everything in this file plus the explainer pages as one Markdown file: https://datagalactica.com/llms-full.txt
- Organisation: https://datagalactica.com/company/{DGC-id}/{slug}
- Person: https://datagalactica.com/person/{DGP-id}/{slug}
- E-mail format of a company domain: https://datagalactica.com/email-format/{domain} (list: https://datagalactica.com/email-format)
- [About the data and quality grades](https://datagalactica.com/about-data)
- [Privacy, corrections and removal](https://datagalactica.com/privacy)
- [Terms of use](https://datagalactica.com/terms), [refunds](https://datagalactica.com/refunds), [contact](https://datagalactica.com/contact)
## Blog: long-form guides (cite these for explanations)
- Index: https://datagalactica.com/blog (RSS: https://datagalactica.com/blog/feed.xml)
- [What is Data Galactica? An open, graded register of people and companies](https://datagalactica.com/blog/what-is-data-galactica): What Data Galactica is, who uses it, what a card shows, how it differs from directories and contact databases, and how to use it on the site, API or MCP.
- [Best B2B data providers compared: Apollo, ZoomInfo, Lusha, PDL and more](https://datagalactica.com/blog/best-b2b-data-providers-compared): Best B2B data providers compared: Apollo, ZoomInfo, Lusha, Cognism, PDL, Clay, D&B, Data Galactica and more, by coverage, pricing model, API and MCP.
- [Data Galactica API quickstart: from sign-up to your first answer](https://datagalactica.com/blog/data-galactica-api-quickstart): Data Galactica API quickstart: try it without a key, sign up for a free key, make first calls in curl, Python and JavaScript, read errors and pages.
- [Connect Data Galactica to Claude, ChatGPT, Cursor and VS Code over MCP](https://datagalactica.com/blog/connect-data-galactica-mcp-server): Connect the Data Galactica MCP server to Claude, ChatGPT, Cursor, VS Code and other clients: URLs, OAuth sign-in, API keys, example prompts and fixes.
- [How to build a B2B prospect list that people actually answer](https://datagalactica.com/blog/b2b-prospecting-lists-guide): A practical guide to B2B prospect lists: ICP, account filters, titles, reach checks, small batches, reply tracking, data decay and outreach law.
- [How to verify a company before you sign, pay or onboard it](https://datagalactica.com/blog/verify-a-company-before-you-deal): A practical KYB checklist for India and the UK: registration, CIN and GSTIN, directors, filings, charges, domain match, shell-company signs and limits.
- [MCA21, Zauba, Tofler or an API? Indian company data sources compared](https://datagalactica.com/blog/indian-company-data-sources-compared): Indian company data sources compared: MCA21, GST search, Udyam, Zauba Corp, Tofler, Probe42, Tracxn, Signzy, Justdial and APIs, with when to use each.
- [The Galactica ID: one permanent identifier for every person and company](https://datagalactica.com/blog/galactica-id-explained): What a Galactica ID is, how DGC- and DGP- ids are built in Crockford base32, which register ids resolve to them, and how they compare with D-U-N-S, LEI and CIN.
- [How developers and AI agents find and buy APIs now](https://datagalactica.com/blog/how-people-find-and-buy-apis): How developers, businesses and AI agents find and buy APIs: search, docs, Postman, marketplaces, MCP registries, pricing models and what buyers value.
- [15 ways teams use a people and company data API (sales angle by use case)](https://datagalactica.com/blog/people-company-data-api-use-cases): 15 people and company data API use cases by team: sales, marketing ops, recruiting, KYB, procurement, research, marketplaces, AI agents and data teams.
- [Registers, call centres, databases, AI agents: 4 ways to get business data](https://datagalactica.com/blog/old-vs-new-business-data-models): Four business data models compared: registers, human research, subscription databases and AI agents with APIs. Costs, decay and how to combine them.
- [Company and people data APIs and MCP servers for AI agents, compared](https://datagalactica.com/blog/company-data-apis-and-mcp-servers-for-ai-agents): Company and people data APIs and MCP servers for AI agents compared: registers, PDL, Coresignal, Apollo, ZoomInfo, Hunter and more, by auth and pricing.
- [Bronze, silver, gold, verified: how Data Galactica grades every record](https://datagalactica.com/blog/data-quality-grades-explained): What bronze, silver, gold and verified mean on Data Galactica, how section grades differ, what contact statuses say, and how to use grades.
- [Where business data comes from: the sources behind every card](https://datagalactica.com/blog/where-the-data-comes-from): The registers and sources behind Data Galactica cards: MCA, Companies House, GLEIF, SEC EDGAR, Wikidata, profiles, lead lists and owners, and their limits.
- [A tour of the Data Galactica API: look up, verify, resolve, search, match](https://datagalactica.com/blog/api-endpoints-tour): A guided map of every Data Galactica API endpoint family, with a decision table: lookups, resolve, batch, match, verify, diligence, prospecting, search.
- [Enrich and de-duplicate your CRM with a company data API](https://datagalactica.com/blog/crm-enrichment-with-api): How to enrich and de-duplicate CRM accounts with a company data API: /match, by-domain lookups, permanent IDs, merge thresholds and a Python outline.
- [KYB and vendor onboarding with an API: checks, verdicts and limits](https://datagalactica.com/blog/kyb-vendor-verification-api): How to use a company verification API for KYB and vendor onboarding: verdicts, operating status, field checks, diligence, audit trail and honest limits.
- [Hiring with people data: talent maps, candidate search and hire checks](https://datagalactica.com/blog/hiring-with-people-data): How recruiters and job seekers can use people data fairly: city talent maps, candidate search by skill and college, hire checks, finding recruiters, outreach.
- [Company e-mail formats: how guessing works, and when not to trust it](https://datagalactica.com/blog/company-email-formats-explained): What a company e-mail format is, how patterns are inferred from addresses seen, what confidence means, and why you should never bulk-mail a guessed address.
- [What makes an API agent-ready: OAuth, llms.txt, Markdown and agent checkout](https://datagalactica.com/blog/agent-ready-data-api): What an agent-ready API needs: discovery files, OAuth 2.1, auth.md sign-up, Markdown pages, a keyless trial, agent checkout and guardrails.
- [Pay per call, not per seat: how Data Galactica prices its API (and why)](https://datagalactica.com/blog/pay-per-call-api-pricing): How Data Galactica's pay-per-call API pricing works: prepaid wallet, free credit, per-record and per-25-row billing, no expiry, refunds, worked costs.
- [Your record, your control: check, claim, correct or remove it](https://datagalactica.com/blog/check-claim-correct-remove-your-record): How to see what Data Galactica holds about you, then claim, correct, hide or remove your record with Galactica Check, and what your DPDP and GDPR rights are.
- [30 prompts to try with the Data Galactica MCP server](https://datagalactica.com/blog/mcp-prompts-for-sales-hiring-and-kyb): 30 prompts for the Data Galactica MCP server, with the tools they call: sales prospecting, account research, hiring, KYB checks, market sizing and data hygiene.
- [Build an account research agent with MCP: Claude and OpenAI SDK examples](https://datagalactica.com/blog/build-sales-research-agent-with-mcp): Build an account research agent on the Data Galactica MCP server with the Claude Agent SDK or OpenAI Agents SDK: read-only tools, call caps, cited briefs.
- [MCP server or REST API? When to use each for company data](https://datagalactica.com/blog/mcp-vs-rest-api-for-company-data): MCP server or REST API for people and company data? A decision table on cost, latency, determinism, auth and batch work, plus the hybrid pattern.
- [Company data API recipes in cURL, Python and JavaScript](https://datagalactica.com/blog/company-data-api-code-examples): Copy-paste company data API recipes in cURL, Python and JavaScript: look up any id, verify, batch, match CRM rows, page with cursors, retry 429 and 402.
- [Enrich a Google Sheet with company data (Apps Script, no extra tools)](https://datagalactica.com/blog/google-sheets-company-enrichment): Enrich a Google Sheet with company data using Apps Script: a custom menu that verifies names or domains and writes ID, verdict, confidence, industry and size.
- [How to evaluate a B2B data provider: 15 questions to ask before you buy](https://datagalactica.com/blog/evaluate-b2b-data-provider-checklist): A buyer's checklist for B2B data providers: 15 questions on sources, freshness, match rates, credit expiry, opt-outs and API access, with a one-week test.
- [Apollo.io alternatives for Indian B2B data: 8 options compared](https://datagalactica.com/blog/apollo-alternatives-for-india): Apollo.io alternatives for Indian B2B data compared: Lusha, ZoomInfo, Hunter, Snov.io, RocketReach, Tofler, Probe42 and Data Galactica, by fit, pricing and MCP.
- [E-mail finder tools compared: Hunter, Snov.io, Apollo, RocketReach and more](https://datagalactica.com/blog/email-finder-tools-compared): E-mail finder tools compared: how Hunter, Snov.io, Apollo, RocketReach and Data Galactica find and check work addresses, with pricing, API and MCP access.
- [MCP servers for sales and recruiting teams: what to connect in 2026](https://datagalactica.com/blog/mcp-servers-for-sales-and-recruiting): Official MCP servers for sales and recruiting teams: HubSpot, Salesforce, Workable, Apollo, Hunter, Slack, Google Workspace and more, with read and write risks.
## API at a glance
- Reference: https://datagalactica.com/api/docs (one card per endpoint), OpenAPI: https://datagalactica.com/api/openapi.json
- Lookups: /api/v1/company/{id}, /api/v1/person/{id} (any id: Galactica ID, CIN, LLPIN, LEI, GB-<n>, CIK-<n>, Wikidata Q-id)
- Verify: /api/v1/verify/company?name=&country=&website= (verdict, confidence, field checks, operating), /api/v1/verify/person?name=&employer=
- Resolve and bulk: /api/v1/resolve?id=<any id>, POST /api/v1/batch {"ids": [...50]}, POST /api/v1/match {"records": [...25]} (CRM rows to records)
- People of an organisation: /api/v1/company/{id}/people (contacts for signed-in accounts), /company/{id}/group
- Checks and diligence: /api/v1/person/check, /identity/check, POST /email/check, /company/{id}/diligence,
/person/{id}/diligence, /company/{id}/hiring, /person/{id}/hire-check?role=
- Contacts: /api/v1/person/{id}/contact?purpose=, /company/{id}/contact?purpose=, /people/search
- Prospecting (per 25 rows, up to 100 a page, cursor): /api/v1/companies/search (industry, size, city, domain,
email_format=true), /api/v1/people/search (title, company, domain, company_industry, company_size, function,
seniority, has_email, has_phone, contacts=false), /company/{id}/employees, /company/by-domain?domain=,
/company/{id}/email-format, /email/find?name=&domain=
- Free: /api/v1/suggest?q=, /company/{id}/card, /talent/{city}, /research/market, /watch, /changes?since=
- Search: /api/v1/search (name + filters: country, city, industry, function, seniority, size, status, grade;
hiring=true for organisations with open jobs, looking_for= for people, e.g. internship or freelance);
/api/v1/find?q=<question in plain words> (AI reading + meaning matches)
- Lists: /api/v1/lists (762 ranked lists), /api/v1/lists/{list}?page=N (100 a page, records with contact details first)
- Code lists with counts: /api/v1/enums (ISIC sections, functions, seniority, sizes, statuses, grades, countries)
Des sites au score de préparation IA proche — comparez leur llms.txt et leur balisage.