Intelligence Center — MCP Developer Documentation
A governed MCP server that turns GlobalData's business intelligence across 20 industry verticals into tools an AI agent can call directly — no bespoke API integration, no SQL, no data export.
A single MCP server endpoint per vertical that exposes structured business intelligence as callable tools. Your AI agent resolves taxonomy, queries live data, and receives structured responses — without custom ETL pipelines or direct database access.
A single endpoint pattern: https://mcp.globaldata.com/{site}/mcp. Disruptor, Technology, Mining, Construction, Power, Oil & Gas, Financial Services, Consumer, and more — each reachable through the same MCP connection with the same tool interface.
The gateway normalises plain-English terms — industries, locations, themes, deal types, company names — to GlobalData's canonical taxonomy before querying, so filters match the catalog and results are accurate.
Any compliant MCP client: Claude Desktop, claude.ai connectors, OpenAI Agents / Responses API, Microsoft Copilot Studio, and custom agents built on the MCP SDK.
Every agent interaction follows a four-stage pipeline from authentication through to structured results.
Authenticate via GlobalData SSO and open the MCP session over Streamable HTTP.
Reveal the tools for the vertical and domain you need. Progressive — tools appear on demand.
Map plain-English terms to canonical taxonomy values with confidence scores.
Run the query with resolved filters. Paginate and chain results by Company ID across domains.
Company ID (returned by list_companies, and by resolve_entities as company_id) is the primary join key across all content domains. Use it for "for this company, give me…" multi-domain queries that chain deals, news, filings, jobs, and patents. A second identifier, cdms_company_id, is also returned alongside it — it's a different ID system used only by a handful of specific tools that ask for it by name; use company_id everywhere else.
search when unsureThe search facade resolves entities, routes to the right tool by intent, and returns data plus delegated_tool, route, and applied_filters. No domain reveal needed before calling it.
The gateway doesn't expose all tools at connect time. A small entry layer is always visible; each domain's tools appear only when you call discover_capabilities. This keeps the agent's tool list focused and avoids overwhelming it with irrelevant capabilities.
Available the moment you connect, no unlock needed.
list_domainssearchresolve_entitiesdiscover_capabilitiesreveal_advancedget_capabilitiestool_searchRevealed per session via discover_capabilities('domain').
list_companieslist_dealslist_newsarticleslist_filinglist_patentslist_jobslist_reports and more…Revealed via reveal_advanced('domain') alongside the standard set.
get_company_analyticsget_deals_analyticsget_news_analyticsget_jobs_analyticsget_filings_analyticsdiscover_capabilities or reveal_advanced emits a notifications/tools/list_changed event — compliant clients re-fetch the tool list automatically. The search facade routes to any tool, revealed or not.search(domain="...", keywords="...") for guaranteed access (parameter renamed from query to keywords — query still works as a legacy alias), or call tools/list again and only call a tool by name once it actually appears in the response. Calling a freshly-revealed tool by name before that can fail with a "tool not found" error — and if the domain isn't in your subscription at all, discover_capabilities/reveal_advanced will now say so directly instead of pretending to reveal it.All authentication flows through GlobalData SSO at login.globaldata.com. Two flows are supported: credentials-based (for server-to-server agents and scripts) and OAuth 2.1 Authorization Code + PKCE (for interactive clients such as claude.ai and Copilot Studio).
Exchange GlobalData username and password for a bearer token. Best for back-end agents, scripts, and server-to-server automation.
POST https://login.globaldata.com/oauth/token Content-Type: application/x-www-form-urlencoded grant_type=password &username=$GD_USERNAME &password=$GD_PASSWORD &scope=openid profile email offline_access
{
"access_token": "eyJ...",
"token_type": "Bearer",
"expires_in": 3600
}Pass the token on every MCP request: Authorization: Bearer eyJ...
Authorization Code flow with PKCE — used by claude.ai connectors, Copilot Studio, and browser-based clients. The user signs in via a browser window; no client secret is stored in the agent.
https://login.globaldata.comhttps://login.globaldata.com/oauth/tokenhttps://login.globaldata.com/.well-known/oauth-authorization-serverWhen using mcp-remote (Claude Desktop), a browser window for GlobalData SSO sign-in opens automatically on first connect.
Replace {site} with your vertical slug (e.g. disruptor, construction, mining). The full slug list is in Section 07.
Edit the Claude Desktop config: Settings → Developer → Edit Config.
{
"mcpServers": {
"GlobalData-MCP": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.globaldata.com/{site}/mcp"
]
}
}
}mcp-remote opens a browser window for GlobalData SSO sign-in on first connect. Subsequent sessions reuse the cached token. Restart Claude Desktop to pick up new tools after a tool set change.
https://mcp.globaldata.com/{site}/mcplist_domains to verify the connection is workingconst response = await client.responses.create({
model: "gpt-4.1",
input: "List GlobalData domains, then find tech deals in APAC.",
tools: [{
type: "mcp",
server_label: "gd_mcp",
server_url: "https://mcp.globaldata.com/disruptor/mcp",
headers: { "Authorization": `Bearer ${process.env.GD_ACCESS_TOKEN}` },
require_approval: "never"
}]
});headers object.https://mcp.globaldata.com/{site}/mcp and a display namecurl -N https://mcp.globaldata.com/disruptor/mcp \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-H "Authorization: Bearer $GD_ACCESS_TOKEN" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'Streamable HTTP transport. POST JSON-RPC 2.0 payloads to /{site}/mcp.
These seven tools are always available the moment you connect — no domain reveal required. They orient your agent, resolve taxonomy terms, route queries, and unlock domain-specific tools.
Lists all available domains for your vertical with their unlock status and the exact command to reveal each one. Your agent's starting point — call this first to understand what data is accessible before deciding which domains to unlock.
Reveals a domain's standard tools for this session. For example, discover_capabilities('companies') makes list_companies and get_company_descriptions callable. Emits notifications/tools/list_changed so compliant clients refresh automatically.
Reveals a domain's analytics tools alongside its standard set. For example, reveal_advanced('companies') additionally exposes get_company_analytics. Use when you need aggregation, trend analysis, or market-level summaries.
Returns the full domain and tool catalog without revealing or enabling anything. Use to inspect what tools exist across all domains before deciding what to unlock — useful for planning a multi-step query sequence.
Searches all tools by keyword across all domains and returns names and summaries even before they are revealed. Use when you know the data type you want (e.g. "analytics", "patent", "project") but aren't sure which domain or tool name to use.
Universal entry point: accepts a natural-language query and optional filters, resolves entities internally, routes to the right tool by intent, and returns data plus delegated_tool, route, and applied_filters. Use when unsure which specific tool fits — dispatches to any tool, revealed or not.
Maps free-text terms (industry, location, theme, deal type, sentiment, and more) to canonical GlobalData taxonomy values with a confidence score (0–1). Call before any direct list_* tool to ensure filters match the catalog. Free-text fields like companyName and keyword pass through unchanged.
20 industry verticals, each with its own endpoint. All share the same content domains (some domains are vertical-specific). Expand a domain to see its tools, or search by tool name.
| Intelligence Center | Notes |
|---|---|
| Consumer Intelligence Center | Adds: pldb domain |
| Retail Intelligence Center | |
| Packaging Intelligence Center | |
| ADS Intelligence Center | Adds: defence_projects domain |
| Financial Services Intelligence Center | |
| Technology Intelligence Center | Adds: ict_contracts domain |
| Disruptor Intelligence Center | Adds: innovation_explorer domain |
| Mining Intelligence Center | Adds: mine_projects, mining_cost_curve domains |
| Oil and Gas Intelligence Center | Adds: upstream, midstream, downstream, oilgas_energy_transition, equipment_and_services, operation_metrics domains |
| Power Intelligence Center | Adds: power_plants, tenders_and_contracts, auctions, td_projects, energy_storage_projects, smart_grid_projects, power_energy_transition, equipment_markets domains |
| Construction Intelligence Center | Adds: contacts, innovation, projects domains |
| Insurance Intelligence Center | Adds: innovation_explorer domain |
| Automotive Intelligence Center | |
| Travel & Tourism Intelligence Center | |
| Foodservice Intelligence Center | |
| Apparel and Footwear Intelligence Center | |
| MEED Projects | Adds: contacts, projects domains |
| Sport Intelligence Center | Adds: sports_athletes, sports_venues, sports_competitions, sports_conferences, sports_deals, sports_brands, stadium_projects, sports_media_revenues domains |
| GlobalData Explorer | |
| Agribusiness Intelligence Center |
fdi_projects and global_ads domains (list_fdiprojects, list_globalads) are available on every vertical above — neither is tied to a single Intelligence Center.Search and list companies by name, industry, type, headquarters country, revenue range, or keyword. Returns structured company records including Company ID — the primary join key used to retrieve related deals, news, filings, jobs, and patents for that company across all content domains.
Retrieve full company profiles for one or more Company IDs: overview text, financials (revenue, net income, EBITDA, employee count), trading information, parent/subsidiary relationships, sector classification, and contact details.
Aggregate company data into buckets for quantitative analysis. Group by industry, country, revenue band, or financial year. Returns counts and metric summaries (revenue, net income, market cap, employee count, growth rates). Requires reveal_advanced('companies').
Search M&A transactions, licensing agreements, partnerships, and collaborations. Filter by deal type, status, rationale, geography, company name, or date range. Returns deal summaries including parties, deal value, and deal date.
Full deal profiles for one or more Deal IDs: description, all parties involved, disclosed financials, deal rationale, advisors, investors, and related company and industry tags.
Aggregate deal data for market-level analysis. Group by deal type, status, year, geography, industry, or theme. Returns counts and total/average deal values. Requires reveal_advanced('deals').
Search curated news articles by company name, industry, theme, news category, sentiment (Positive / Negative / Neutral), location, or date range. Returns article summaries with sentiment scores and tagged entities.
Full article content for one or more article IDs: headline, body text, publication date, source, sentiment score, mentioned companies, industries, and themes.
Aggregate news volume and sentiment over time. Group by company, industry, month, or theme to surface coverage trends and sentiment shifts. Requires reveal_advanced('news').
Search regulatory and corporate filings — annual reports, quarterly reports, earnings call transcripts, investor presentations — by company, filing type, sector, country, or date.
Extract specific sentences or passages from a filing document by keyword or topic. Ideal for surfacing ESG disclosures, risk factors, or forward-looking statements without reading the full document.
Returns the full structured filing document broken into labelled sections, making it easy to navigate directly to a specific part of the filing.
Returns the raw full text of a filing. Documents are stored in cloud storage; a time-limited access URL is returned for download or further processing.
Aggregate filing volume and sentiment by filing type, sector, country, or year. Requires reveal_advanced('filings').
Search patents by IPC classification code, application year, assignee company, inventor country, or keyword. Returns patent summaries including application number, title, assignee, and filing and grant dates.
Full patent record for one or more patent IDs: abstract, full IPC classification tree, inventor list, applicant/assignee details, filing date, and grant status.
Search job postings by title, occupation category, seniority level, company name, country, state/city, or posting date. Returns postings with salary ranges (local currency and USD-normalised) and required education/skills.
Full job posting content for one or more job IDs: full description text, salary min/max, education requirements, occupation classification, and skills tags.
Aggregate job posting data for hiring trend analysis. Group by occupation, country, company, education type, or date. Returns counts and salary analytics (min, max, median). Requires reveal_advanced('jobs').
Fetches multiple jobs analytics groupings simultaneously (e.g. occupation + location + education) in one call, reducing round trips for dashboard-style queries. Requires reveal_advanced('jobs').
Search decision-maker and senior professional contacts at tracked companies. Filter by name, job title, department, company name, or country. Returns contact profiles with verified business contact details.
Search GlobalData analyst research reports including market forecasts, sector analyses, country intelligence reports, and thematic studies. Filter by topic, industry, country, or publication date.
Full report details for one or more report IDs: abstract, key findings summary, scope, methodology overview, lead analyst, and publication date.
Search social media posts that mention tracked companies, industries, or themes. Filter by sentiment, company, industry, or date range. Returns post summaries with sentiment scores and entity tags.
Retrieve actionable sales lead records: BD-relevant events classified by supplier relevance. Sources include deal announcements, hiring signals, regulatory events, and news triggers. Filter by lead type, industry, company, or date.
Aggregate buying signal volume by signal/lead type, company, or month. Requires reveal_advanced('buying_signals').
Search innovation records in the construction and infrastructure sector: new technologies, start-up activity, R&D initiatives, and disruptive products. Filter by innovation type, company, theme, or country.
Retrieve quantitative market data: sector sizing, market share, CAGR, growth forecasts, and production capacity metrics. Calls GlobalData's backend structured data API which performs its own entity matching internally — resolve_entities is not required before calling this tool. Pass the user's natural language question directly as the query parameter. Unlock via discover_capabilities('market_data'). Subscription entitlement required — contact DDStechSupport@globaldata.com if access is denied.
Search mining projects by commodity type, mine method, development stage, country, or company. Returns mine project records including resource estimates, production status, owner/operator, and geographic location.
Full detail record for a single mine project by mineId: description, project value, timeline, equity partners, operators, involved companies, location, and annual production data by commodity/year.
Search infrastructure and capital projects — energy, construction, transportation, utilities — by stage (Announced, Planning, Execution, Completed), sector, value, or geography. Filter by construction start, contract award, tender issue/submission, or project start/end date ranges, company role, development likelihood, or event status.
Full project detail records: description, value breakdown, timeline, involved companies and their roles, location, and status.
Aggregate project counts and investment value by status, industry, country, or start date. Requires reveal_advanced('projects').
projects domain above.Search and list Foreign Direct Investment (FDI) projects worldwide — cross-border capital investment tracking by investing company, destination country, sector, or theme. Subscription entitlement required.
List IT software, service, or telecom contract records across the technology market, selected via a category parameter (Software / Service / Telecom). Does not require a company ID — filterable by date and category.
Full ICT contract details for a specific company: vendor, client, contract value, duration, and category. Requires a category value.
Aggregate contract volume and value by vendor, client, industry, contract type, region, or year. Requires reveal_advanced('ict_contracts').
Search Product Lifecycle Database (PLDB) records worldwide: products by industry, location, closure material, closure type, or pack sub-type. Returns product summaries (title, industry, location, published date).
Search TV, print, and YouTube advertising records for a company — media type, brand/product, country, and YouTube engagement metrics (views, likes).
Full ad record detail by ID.
Search aerospace, defence, and security project records by sector, stage, value, or geography.
Full defence project detail: contractor, client, project value, and description.
Search emerging-technology and disruptor innovation profiles by title, location, or industry taxonomy.
Full innovation profile detail: overview, taxonomy, and related companies.
Search power generation plants across every technology type — thermal, nuclear, hydro, wind, solar, geothermal, ocean, biopower — by technology, location, status, capacity, or company role (owner, developer, EPC, manufacturer).
Full power plant detail by ID.
Identifies ageing plants at decommissioning risk.
Announced and under-construction plant pipeline.
mine_projects, not the same domain.Search mining production cost curve records by commodity, cost curve type, product type, mine, or year.
Cross-mine/company/country production cost benchmarking, with cumulative production and chart-ready cost bars. Requires reveal_advanced('mining_cost_curve').
Search athletes/players by sport discipline, country, gender, or active/retired status.
Search venues and stadiums by sport discipline, country, or operational status.
list_properties → list_competitions (by propertyIds) → list_calendarevents (by competitionId or propertyId).Search competition-series ("properties", e.g. "FIM Sidecar World Championship") by sport, country, year, competition type, or gender.
Search competitions/tournaments/leagues by sport, country, active status, or year.
List calendar events (fixtures, rounds, matches) for a specific competition or property-series.
Search sport conferences and industry events by country/city, conference type, status, or date range.
deals domain instead.Sponsorship agreements between brands and rights-holders, by sport, geography, theme, status, value, or date.
Media and broadcast rights deals.
Host-city / event hosting-rights bidding deals.
Financial/M&A deals involving sport entities.
Aggregate sponsorship deal volume/value. Requires reveal_advanced('sports_deals').
Aggregate media/broadcast rights deal volume/value.
Aggregate host-city/event bidding deal volume.
Aggregate financial/M&A sport deal volume/value.
Search and rank sponsorship brands by industry, HQ, sport sponsored, beneficiary, relationship, status, or total/annual spend.
Stadium Construction Database — search stadium/arena construction and redevelopment projects by geography, stage, value, contractor, or funding.
Aggregate stadium project counts/value by country, stage, sector, contractor, or year. Requires reveal_advanced('stadium_projects').
list_media_revenues discovers which revenue series exist for a property, then compare_media_revenues returns year-by-year figures.Discover which media-rights revenue series exist for a sport property.
Year-by-year media-rights revenue figures (US$m) for comparison across properties or years.
og_* tool-name prefix; all SQL-backed (not Elasticsearch) against the Oil & Gas database — the first production domain to use this pattern.Exploration blocks, fields, and wells — og_blocks_listing/details, og_fields_listing/details, og_international_wells_listing/details, og_us_wells_list/details.
Gas processing, gas & liquid storage, LNG, pipelines — og_gas_processing_listings/details, og_gas_storage_listing/details, og_liquid_storage_listing/details + advanced og_liquid_storage_analytics, og_lng_liquefaction_listing/details, og_lng_regasification_listing/details, og_pipelines_listing/pipeline_details.
Petrochemical plants and refineries — og_petrochemicals_listing/petrochemical_details, og_refinery_listing/details.
Carbon capture (CCUS) and O&G hydrogen production — og_carbon_capture_listing/details, og_hydrogen_plants_listing/details (a separate implementation from Power's own hydrogen-plants tool — each cross-references the other to prevent misrouting).
Services/equipment/construction contracts and upstream/midstream project records — og_contracts_listing/contract_details, og_projects_listing/project_details.
Company production/reserves/lifting-cost KPIs — og_operational_metrics.
power_plants domain above. All Elasticsearch-backed except equipment_markets (SQL-backed).list_power_tenders / get_power_tenders_descriptions — power-sector tenders and contracts by category, segment, equipment, technology, location, or client/vendor.
list_power_auctions / get_power_auction_descriptions — renewable-energy auction results, awarded capacity, clearing prices/tariffs.
list_power_tnd_projects / get_power_tnd_projects_descriptions — transmission & distribution line/substation projects by type, location, status, voltage, or system operator.
list_power_storage / get_power_storage_descriptions — battery, pumped hydro, and hydrogen storage projects by technology, location, capacity, or company role.
list_smart_grid_projects — AMI, DSM, EV charging, grid modernization, and microgrid projects.
list_hydrogen_plants / get_hydrogen_plant_descriptions — power-sector hydrogen plants by operator, process, technology, or development stage.
list_equipment_markets — power equipment market value/volume/capacity by technology, region, or year.
A nine-step lifecycle covers every MCP session from connection through to presenting results.
list_domains to see available domains for your vertical with unlock status. Skip when your agent already knows the tool it needs.discover_capabilities('domain') exposes standard tools; reveal_advanced('domain') adds analytics. Or skip revealing entirely and use the search facade — it dispatches to any tool, revealed or not. Immediately after a reveal call, prefer search(domain, keywords) over calling the new tool by name — your client's local tool list may not have refreshed yet.resolve_entities first with a domain hint. Free-text fields (companyName, keyword) pass through unchanged and do not need resolving.list_* tool with resolved filters, or use search by natural-language intent. At least one selective filter is required — a bare geography with no other filter is rejected pre-flight.applied_filters (true scope after canonicalisation), resolved (term mapping with confidence scores), dropped_filters and warnings (anything silently lost). Counts are meaningless until scope is confirmed.pagination.has_more is set, re-call with the next page_number. The envelope provides next_page.example_args ready to use directly.industry · location · theme · hqCountrydealType · dealStatus · dealRationale · acquirerCountrynewsCategory · sentimentcommodity · mineMethod · mineStagecompanyName · keyword · job title · text search fields
dealType/sentiment/newsCategory take a separate, simpler path: an exact/alias match against a known list first, and — new as of July 2026 — an AI-assisted fuzzy match as a second attempt before a value is dropped (so a near-miss like "Geographical Expansion" now still resolves to "Geographic Expansion" instead of silently failing). Either way, check the resolved/metadata block for how a value was matched before trusting it.| Field | Meaning |
|---|---|
result | The data payload — actual records from the backend tool. |
delegated_tool · route | Which tool the facade ran and why. Confirm intent was interpreted correctly. |
applied_filters | True scope after canonicalisation. Always verify before reporting counts or quoting results. |
resolved | Full resolve_entities payload: each input term, canonical value mapped to, and confidence score. |
dropped_filters · warnings | Any filter that couldn't be honored or was silently dropped. |
scope_warning | Raised when a primary filter couldn't apply and the call continued with degraded scope. |
pagination | page, page_size, returned, total_records, total_pages, has_more, next_page.example_args |
routing_hint · no_results_hint | Sibling tools to try when routing was ambiguous or results are empty. |
unknown_fields | Fields passed to fields= that didn't match — the rest still project correctly. |
result_set_too_large error that lists the filters which would make it valid.Typical call sequences per domain. Each example shows the intent and the tool calls in order.
resolve_entities — hqCountry="US", industry="Technology" → canonical valueslist_companies — pass resolved hqCountry, industrylist_companies — companyName="Siemens" → extract Company_IDget_company_descriptions — pass Company_IDreveal_advanced('companies')resolve_entities — industry="Mining"get_company_analytics — groupBy="industry", resolved industryresolve_entities — dealType="Acquisition", location="Asia-Pacific", industry="Technology"list_deals — resolved filters + from_date for current yearlist_deals — companyName="X" → extract Deal_IDsget_deals_descriptions — pass Deal_IDsreveal_advanced('deals')get_deals_analytics — groupBy="dealType", date rangeresolve_entities — sentiment="Negative"list_newsarticles — companyName=X, resolved sentiment, date rangereveal_advanced('news')resolve_entities — industry="Renewable Energy"get_news_analytics — groupBy="month", resolved industry, date rangelist_filing — companyName=X, filingType="Annual Report"list_filing — find filing, extract Filing_IDget_filingsentences — Filing_ID, keyword="supply chain risk"list_filing → Filing_IDget_filing_document_text — returns time-limited download URLlist_patents — assignee="Company X", ipcCode="H01M"resolve_entities — industry="Battery Technology"list_patents — resolved industry as keyword filterresolve_entities — location="Germany"list_jobs — companyName=X, occupation="Engineering", resolved locationreveal_advanced('jobs')resolve_entities — industry="Technology"get_jobs_analytics — groupBy="occupation", resolved industrylist_contacts — companyName="X", seniority="C-Level"resolve_entities — industry="Automotive", theme="Electric Vehicles"list_reports — resolved filterslist_reports → extract Report_IDget_report_descriptions — pass Report_IDresolve_entities — sentiment="Negative" (if filtering by sentiment)list_socialmediaposts — companyName=X, from_date this week, resolved sentimentresolve_entities — industry="Pharmaceuticals"list_buying_signals — resolved industryreveal_advanced('buying_signals')get_buying_signals_analytics — groupBy="month", date rangeresolve_entities — theme="Sustainability"list_innovations — resolved themeresolve_entities — industry="Energy", location="Saudi Arabia"list_projects — resolved filters, stage="Announced", date rangereveal_advanced('projects')resolve_entities — location="Middle East"get_projects_analytics — groupBy="status", resolved locationresolve_entities — location="Africa"list_fdiprojects — companyName="China"-scoped keyword, resolved destination locationlist_ict_contracts — category="Service"list_ict_contracts → extract contract IDget_ict_contract_descriptions — contract ID, category="Software"reveal_advanced('ict_contracts')get_ict_contract_analytics — groupBy="vendor", date rangeresolve_entities — location="Germany", pldb_pack_sub_type="Bottle - Glass"list_pldb — resolved location, pack sub-typeSource types, coverage scope, and important caveats per tool. Coverage summary = directly documented scope. Inherits parent = shares the primary tool's coverage.
| Domain | Tool | Status | Sources | Coverage | Watch for |
|---|---|---|---|---|---|
| companies | list_companies | Coverage summary | Company registries, financial data providers, GlobalData research | Company profiles globally across all tracked organisations | Company ID is the master join key for all domains |
| companies | get_company_descriptions | Inherits parent | — | Detailed financials, subsidiaries, sector classification | Inherits list_companies coverage |
| companies | get_company_analytics | Inherits parent | — | Aggregated metrics by industry, country, revenue band, year | Requires reveal_advanced; inherits list_companies coverage |
| deals | list_deals | Coverage summary | Company announcements, press releases, financial news, transaction databases | M&A, licensing, partnerships, JVs across all tracked companies | Deal value may be undisclosed; some deals announced after close |
| deals | get_deals_analytics | Inherits parent | — | Aggregated counts and deal values | Inherits list_deals coverage |
| news | list_newsarticles | Coverage summary | Licensed news feeds, press releases, trade publications, analyst commentary | Curated articles on GlobalData-tracked companies, industries, and themes | Sentiment score is model-derived — treat as a signal, not a definitive label |
| filings | list_filing | Coverage summary | Regulatory filing portals, stock exchange disclosures, company investor relations | Annual/quarterly reports, earnings transcripts, investor presentations | Full document text is retrieved via a time-limited access URL |
| filings | get_filingsentences | Inherits parent | — | Sentence-level extraction with keyword matching | Inherits list_filing coverage |
| patents | list_patents | Coverage summary | International and national patent offices and publication databases | Patents across technology, energy, materials, and other sectors | For drug-level patent expiry data, use the Healthcare Pharma MCP |
| jobs | list_jobs | Coverage summary | Job board aggregators and company career pages | Postings from GlobalData-tracked companies; salary normalised to USD where available | Historical postings may be deduplicated; not all postings include salary data |
| jobs | get_jobs_analytics | Inherits parent | — | Hiring trends by occupation, location, or company | Salary analytics require sufficient postings per bucket for reliability |
| contacts | list_contacts | Coverage summary | Company websites, professional directories, publicly available business profiles | Senior decision-makers and professionals at tracked companies | Requires contacts entitlement; construction and meedprojects only |
| reports | list_reports | Coverage summary | GlobalData analyst research team publications | Market sizing, forecasts, sector analyses, and country intelligence reports | Report availability varies by subscription tier |
| social | list_socialmediaposts | Coverage summary | Social media platforms via licensed data partnerships | Posts mentioning tracked companies and themes, with sentiment scores | Platform coverage and recency depend on licensing agreements |
| buying_signals | list_buying_signals | Coverage summary | Aggregated from deals, news, hiring, and regulatory event data | BD-relevant events classified by supplier relevance | Renamed from "leads"; classification is rule-based, verify individually before outreach |
| buying_signals | get_buying_signals_analytics | Inherits parent | — | Aggregated by signal/lead type, company, or month | Inherits list_buying_signals coverage |
| innovation | list_innovations | Coverage summary | Construction sector R&D databases, start-up intelligence sources | New technologies and innovation activity in the construction sector | construction vertical only |
| mine_projects | list_mineprojects | Coverage summary | Mining project databases, company announcements, regulatory filings | Active and planned mining projects globally with resource and production data | mining vertical only |
| mine_projects | get_mine_descriptions | Inherits parent | — | Full mine detail: equity partners, operators, involved companies, annual production by commodity/year | Takes a single mineId, not a comma-separated list; inherits list_mineprojects coverage |
| market_data | get_market_data_analytics | Coverage summary | GlobalData structured data and market intelligence databases | Sector-level numeric market data: sizing, share, growth rates, CAGR forecasts — available on all verticals | Subscription entitlement required; resolve_entities not needed — pass natural language query directly |
| projects | list_projects | Coverage summary | Project announcements, tender/contract award notices, company disclosures | Infrastructure and capital projects — energy, construction, transportation, utilities | construction and meedprojects verticals only |
| projects | get_project_descriptions / get_projects_analytics | Inherits parent | — | Full detail records and aggregated counts/values | Inherits list_projects coverage |
| fdi_projects | list_fdiprojects | Coverage summary | Cross-border investment announcements, company disclosures, regulatory sources | Foreign Direct Investment projects globally — investing company, destination country/sector | Subscription entitlement required; available on all verticals |
| ict_contracts | list_ict_contracts / get_ict_contract_descriptions | Coverage summary | Vendor/client contract disclosures, technology market research | IT software, service, and telecom contracts — vendor, client, contract value, duration, category | technology vertical only; subscription entitlement required |
| ict_contracts | get_ict_contract_analytics | Inherits parent | — | Aggregated by vendor, client, industry, contract type, region, or year | Inherits list_ict_contracts coverage |
| pldb | list_pldb | Coverage summary | Product Lifecycle Database — retail and packaging tracking sources | Products worldwide by industry, location, closure type/material, pack sub-type | consumer vertical only |
| global_ads | list_globalads / get_globalads_descriptions | Coverage summary | TV, print, and YouTube ad monitoring sources | Advertising records for a company — media type, brand/product, country, engagement metrics | available on all verticals |
| defence_projects | list_defenceprojects / get_defenceproject_descriptions | Coverage summary | Defence procurement announcements, contractor disclosures | Aerospace, defence, and security project records — sector, stage, value, geography | ads vertical only |
| innovation_explorer | list_innovationexplorer / get_innovationexplorer_descriptions | Coverage summary | Emerging-technology and start-up intelligence sources | Disruptor/emerging-tech and Insurance innovation profiles — title, location, taxonomy, related companies | disruptor, insurance verticals |
| power_plants | list_plants / get_plant_descriptions | Coverage summary | Power generation asset databases, company disclosures, regulatory filings | Power plants across every technology type — thermal, nuclear, hydro, wind, solar, geothermal, ocean, biopower | power vertical only |
| mining_cost_curve | list_mining_cost_curve | Coverage summary | Mining production cost databases, company disclosures | Production cost curve line items by commodity, mine, and year, for cross-industry benchmarking | mining vertical only; sibling to mine_projects, not the same domain |
| mining_cost_curve | get_mining_cost_curve_analytics | Inherits parent | — | Cross-mine/company/country cost benchmarking with cumulative production | Inherits list_mining_cost_curve coverage |
| sports_athletes | list_athletes | Coverage summary | Sport industry databases, federation/league sources | Athletes/players by sport discipline, country, gender, active status | sport vertical only |
| sports_venues | list_venues | Coverage summary | Sport industry databases, venue operator sources | Venues/stadiums by sport discipline, country, operational status | sport vertical only |
| sports_competitions | list_properties / list_competitions / list_calendarevents | Coverage summary | Sport industry databases, federation/league sources | Competition-series, competitions/tournaments, and their calendar events (fixtures, rounds) | sport vertical only; three-level hierarchy, see domain explorer above |
| sports_conferences | list_conferences | Coverage summary | Sport industry event calendars | Sport conferences and industry events by country/city, type, status, date | sport vertical only |
| sports_deals | list_sponsorship_deals / list_media_deals / list_bidding_deals / list_financial_sport_deals | Coverage summary | Sport industry deal announcements, company disclosures | Sponsorship, media-rights, host-city bidding, and financial/M&A deals in sport | sport vertical only; one tool per deal kind |
| sports_brands | list_sponsorshipbrands | Coverage summary | Sport sponsorship tracking sources | Sponsorship brands ranked by industry, HQ, sport, relationship, spend | sport vertical only |
| stadium_projects | list_stadium_projects / get_stadium_projects_analytics | Coverage summary | Stadium Construction Database, contractor disclosures | Stadium/arena construction and redevelopment projects; aggregated by country, stage, sector, contractor, year | sport vertical only |
| sports_media_revenues | list_media_revenues / compare_media_revenues | Coverage summary | Sport media-rights market data sources | Media-rights revenue series per property; year-by-year US$m comparison | sport vertical only |
| upstream / midstream / downstream / oilgas_energy_transition / equipment_and_services / operation_metrics | og_* (~25 tools — see domain explorer above) | Coverage summary | Oil & Gas project databases, company disclosures, regulatory filings | Exploration blocks/fields/wells, gas & liquid storage, LNG, pipelines, refineries, petrochemicals, carbon capture, hydrogen, contracts, and operational KPIs | oil and gas vertical only; SQL-backed, not Elasticsearch — see developer bible for the pattern |
| tenders_and_contracts / auctions / td_projects / energy_storage_projects / smart_grid_projects / power_energy_transition / equipment_markets | list_power_* / list_smart_grid_projects / list_hydrogen_plants / list_equipment_markets (see domain explorer above) | Coverage summary | Power-sector tender/auction registries, project databases, company disclosures | Power tenders, renewable auctions, T&D projects, energy storage, smart grid, power-sector hydrogen, equipment markets | power vertical only |
How the platform handles tool changes and how your client picks them up.
Every fresh MCP session receives the latest tool definitions at initialize / tools/list. There is no cross-session caching of tool schemas on the client.
When the tool set changes, the gateway emits notifications/tools/list_changed. Compliant MCP clients re-fetch the tool list automatically without requiring a reconnection.
search immediatelyThe search facade routes to newly shipped tools on deployment — even before a client has been restarted or revealed the domain.
| Client | How to pick up new tools |
|---|---|
| Any fresh connect | Always automatic |
| OpenAI Responses API | Automatic — tool list fetched per run |
| claude.ai connector | Usually automatic via notifications/tools/list_changed; toggle the connector off and on if needed |
| Claude Desktop | Manual — restart the app to reload tool definitions |
| Copilot Studio | Manual — re-sync the MCP tool and republish the agent |
| Custom MCP client | Depends on whether the client handles notifications/tools/list_changed; otherwise reconnect |
https://mcp.globaldata.com/{site}/mcp. The same endpoint handles both standard request/response and streaming (SSE) responses.discover_capabilities('companies') (or any domain name) to reveal that domain's standard tools. Call reveal_advanced('companies') to also reveal analytics tools. Which tools you can actually see depends mainly on your subscription's entitlement — a domain your subscription doesn't include won't reveal no matter how many times you call it, and the response will say so directly. For an entitled domain, the reveal happens server-side immediately, but your client may need a moment (or a tools/list re-fetch) to see the new tools — see fq15.search facade — it routes to any tool, revealed or not, and doesn't depend on your client's local tool list being current. Calling a specific list_* or get_* tool by name does require the domain to be revealed and your client to have picked up the updated tool list — which is why search is the more reliable choice right after a reveal call.consumer, retail, packaging, ads, financialservices, technology, disruptor, mining, oilgas, power, construction, insurance, automotive, tourism, foodservice, apparel, meedprojects, sport, explorer, agri.projects and contacts domains are only on construction (63) and meedprojects (71). The innovation domain is construction only. The mine_projects domain is mining (58) only. The ict_contracts domain is technology (55) only. The pldb domain is consumer (1) only. The separate fdi_projects domain (list_fdiprojects) is the exception — it's available on every vertical, not just construction/meedprojects.resolved block — low-confidence mapping may have searched the wrong scope; (2) no_results_hint — suggests sibling tools that may have the data; (3) ensure at least one selective filter was passed — bare geography alone is pre-flight rejected.applied_filters — this is the true scope after canonicalisation. Also check dropped_filters and warnings for filters that were silently not applied.dealType or sentiment) work a little differently: a plain text match is tried first, and an AI-assisted fuzzy match is tried second before giving up — you can tell which happened from the match method in the resolved block. Always inspect that block before trusting results.list_companies) as arrays to the next domain tool. Company ID is the primary join key across all content domains. For example, pass company_ids=[12345] to list_deals or list_filing.pagination.has_more. If true, use pagination.next_page.example_args from the response for the exact arguments to pass in your next call — simply increment page_number by 1.industry, location, theme, hqCountry. Deal-specific: dealType, dealStatus, dealRationale. News: newsCategory, sentiment. Mining: commodity, mineMethod, mineStage. Free-text fields (companyName, keyword) pass through unchanged.list_contacts requires a contacts entitlement. Contact the DDS team at Contact us if you receive an entitlement error.not_entitled response first — if your subscription doesn't include the domain, no amount of re-revealing or reconnecting will make the tool appear. Beyond that, it's a client tool-list refresh issue: the tool is enabled server-side, but your client's local list hasn't caught up yet. For a domain you just revealed with discover_capabilities/reveal_advanced, call search(domain, keywords) instead of the tool by name — it doesn't depend on your local list at all. If you need the tool callable by name, re-fetch tools/list and confirm it appears before calling it. Note that if your client doesn't maintain a persistent connection to the gateway, it may need to re-discover from scratch on every call rather than the reveal "sticking" — this is expected, not a bug. For persistently missing tools after a platform update: reconnect or restart your client — Claude Desktop requires a full app restart, claude.ai needs the connector toggled off then on, and Copilot Studio needs the MCP tool re-synced and republished.resolved block, and the applied_filters from the response to Contact us.Getting connected and resolving issues.
Request GlobalData MCP credentials or register an OAuth redirect URI by contacting us. Include your client type (Claude Desktop, claude.ai, OpenAI, Copilot Studio, or custom), the vertical(s) you need, and your preferred authentication flow.
After connecting, call list_domains. A clean domain list with unlock status for each domain confirms that authentication, transport, and progressive discovery are all working end to end. If the call fails, check your bearer token and ensure your subscription covers the vertical slug you connected to.
| Endpoint | Description | Auth |
|---|---|---|
/{site}/mcp | MCP protocol endpoint (one per vertical) | SSO required |
/.well-known/oauth-authorization-server | OAuth 2.1 server metadata | Public |
/authorize | OAuth authorization endpoint | Public |
/token | Token exchange endpoint | Public |
/health | Service health check | Public |
/.well-known/ai-catalog.json | Agentic Resource Discovery manifest — machine-readable list of every vertical's MCP server, for AI crawlers that don't already know the site slugs | Public |