Platform Evaluation Guide

This guide is for teams evaluating World Monitor. For the specialized variants and what each one monitors, see the Platform Overview. This guide gives procurement, security, and engineering teams a concise view of the current platform.

Technical capabilities

World Monitor combines an interactive 3D globe and flat map, configurable intelligence panels, source-attributed news and OSINT, market and economic data, country risk and resilience scoring, infrastructure and supply-chain analysis, scenario modelling, AI-generated briefs, and alerting workflows. The generated product inventory records a shared map-layer catalog, concrete panel implementations, and specialized product variants. Panels extend the shared Panel base in src/components; the base class itself is excluded, and a class reused by multiple variants is represented once.

Capability	Current scope
Country risk and resilience	Country Instability Index (CII v14) scores with signed 24-hour movement for 31 Tier-1 countries. The Country Resilience Index scores a 196-country public universe and ranks 169 of 196 countries using 72 indicators across 21 active dimensions and 6 domains on a 6-hour schedule.
Signal and infrastructure analysis	Geographic convergence, temporal anomalies, and cross-source correlation combine related events. Infrastructure Cascade models dependencies across a 350-node infrastructure graph.
Decision support	Source-linked AI briefs, probabilistic forecasts with machine-checkable resolution criteria and scorecard context, Route Explorer, and the Scenario Engine.
Company and procurement workflows	The MCP company-intelligence tool combines market context, SEC filing search, and material 8-K events. Global Procurement Intelligence uses seeded public opportunities from supported government and multilateral portals. Availability still depends on deployment credentials and source health.
Delivery surfaces	Hosted web dashboard and installable PWA, native desktop applications for macOS, Windows, and Linux, and a self-hosted stack. The desktop application includes a local Node.js sidecar and supports local LLM endpoints such as Ollama and LM Studio.
Language coverage	The interface has localized and right-to-left layouts. Arabic and Persian use right-to-left layouts. This documentation site is currently available in English and Simplified Chinese.

For the complete product surface, see Features & Interface. For the processing model, caching, resilience controls, and deployment components, see Architecture. Methodology pages explain how the platform calculates algorithms and composite signals.

Update frequency and freshness

World Monitor does not apply one refresh interval to every dataset. Each source follows a cadence that matches how frequently its upstream provider publishes useful observations. Fast-moving feeds use live or minute-level pipelines; institutional datasets retain their daily, weekly, or monthly publication schedules. The following cycles describe specific product surfaces. REST and MCP freshness budgets can differ from dashboard polling intervals.

Data domain	Normal product cycle	Important qualification
Maritime tracking	AIS uses a live WebSocket when enabled; snapshot surfaces refresh in about 5–15 minutes	MCP live-transit summaries use a 30-minute freshness budget. PortWatch-derived products are slower, with 12–36-hour budgets and source values that can be weekly.
Aviation	Live military-flight and airport-operations surfaces refresh in about 2–5 minutes	Other flight feeds are provider-dependent; MCP flight-backed tools use a 30-minute freshness budget.
Market data	The dashboard refreshes market quotes every 12 minutes	MCP market data uses a 30-minute freshness budget. CFTC, FRED, and other institutional series follow their own publication schedules.
Dashboard news and OSINT	20-minute dashboard refresh cycle	Publisher timing varies. A refresh can return the same latest article when no new item is published.
Conflict and unrest	GDELT bulk materialisation runs on a 15-minute cycle	The MCP unrest slice permits data up to 2 hours old before it is marked stale. Other conflict sources keep source-specific schedules.
Country Resilience Index	Scheduled every 6 hours	A successful cycle does not imply that every upstream series published a new value.
Institutional economic data	Daily, weekly, or monthly	BIS, Eurostat, IMF, World Bank, and similar datasets retain their native release calendars.

These are operating targets, not a promise that an upstream publisher will release a new observation in each window. Cache-backed API and MCP responses include timestamps and a stale indicator where supported. See the MCP tool reference for per-tool freshness budgets, Health endpoints for monitoring semantics, and status.worldmonitor.app for current service status.

Sources and provenance

World Monitor publishes a public, searchable Data Source Catalog. Its manifest-derived inventory lists active providers, upstream hosts, structured endpoints, and news and OSINT feeds. The application feed registry and Telegram catalog are separate inventory layers and should not be added together. Across those inventories, World Monitor combines structured APIs, public and licensed datasets, official government and multilateral sources, news feeds, and reviewed OSINT. Representative providers include ACLED, UCDP, GDELT, UNHCR/OCHA, IMF, World Bank, BIS, Eurostat, FRED, USGS, NASA, Open-Meteo, Cloudflare Radar, AIS and ADS-B providers, and hundreds of news publishers. RSS configuration can include source tier, state affiliation, and propaganda-risk metadata. Telegram configuration uses a different schema with tier, topic, and region metadata. Cross-source corroboration can improve ranking and confidence, but it is not a universal prerequisite for every alert. Start with the public Data Source Catalog to search and filter providers by signal domain, source type, country of origin, or country covered. Origin is the publisher’s home country; coverage is the geography a feed reports on. For collection methods, coverage, credibility tiers, and bias metadata, use the Data Sources methodology. For the host-level attribution and terms-review posture, use the machine-generated Source Attribution inventory. Generation checks detect repository drift, but an inventory entry is not by itself a legal clearance. Source availability and licensing can differ by feature and deployment model.

Integration options and requirements

Option	Best for	Main requirements
Hosted dashboard	Analysts who need the complete visual product	A modern browser. Public capabilities need no customer-side integration; paid features require the applicable World Monitor account or plan.
Live map and panel embed	Articles, portals, and partner products	A page that permits an HTTPS iframe or script tag. The public live map needs no key. Other allowlisted panels check the embedding account’s API key, not the page visitor. Tweet bodies are never embedded.
REST API	Applications, data pipelines, and higher-volume batch reads	Generated service contracts and versioned HTTP paths described by OpenAPI 3.1. Standard server-to-server access uses an API key in `X-WorldMonitor-Key`. See the reference for current routing status, then review the quickstart, authentication, versioning policy, and REST rate limits.
MCP server	Claude, Cursor, and custom AI agents	Streamable HTTP with a live tool registry, workflow prompts, interactive apps, and optional JMESPath response filtering. Catalog discovery is public; data calls require OAuth 2.1 or a World Monitor API key. MCP entitlements and metering are documented separately from REST API-plan allowances.
Official SDKs and CLI	Python, Ruby, Go, and JavaScript applications or shell automation	The relevant language runtime and an API key for data calls. The zero-dependency SDKs and npm CLI support hosted and self-hosted base URLs.
Notification channels	General domain alerts to operational systems	A billed PRO account and Clerk bearer token. Supported delivery channels include signed HTTPS webhooks, Telegram, Slack, Discord, email, and Web Push. Availability depends on the event domain and channel configuration.
Shipping webhooks	Chokepoint disruption notifications	A PRO-capable API key and a public HTTPS callback that passes the documented SSRF checks. Deliveries are HMAC-SHA256 signed and at least once. These webhooks are domain-specific, not a general platform-wide event bus.
Agent discovery	Automated onboarding and agent-native workflows	Standard `.well-known` metadata, `llms.txt`, OpenAPI assets, an MCP server card, and an installable agent-skill catalog. Discovery metadata is public; data-bearing operations still require the applicable credentials.
Self-hosted deployment	Teams that need to operate the stack themselves	Node.js 22+, Docker or Podman, Redis, deployment secrets, and compliance with the AGPL or separate commercial license. See Getting Started.

For browser-based API calls, review the CORS policy. For generated clients, the bundled OpenAPI specification is the machine-readable contract.

Getting Started

Usage

Guides

Company & Legal

Platform Evaluation Guide

Technical capabilities

Update frequency and freshness

Sources and provenance

Integration options and requirements

​Technical capabilities

​Update frequency and freshness

​Sources and provenance

​Integration options and requirements

Technical capabilities

Update frequency and freshness

Sources and provenance

Integration options and requirements