{"id":"marc-shade-world-intel-mcp","name":"world-intel-mcp","af_score":66.2,"security_score":24.8,"reliability_score":35.0,"what_it_does":"world-intel-mcp is an MCP (Model Context Protocol) server and companion CLI/dashboard/collector for pulling and analyzing publicly available global intelligence across 30+ domains (markets, FX, bonds, SEC filings, conflict/security, military/aviation, geospatial datasets, news, cyber threats, etc.). It exposes many MCP tools (100+ per README) and supports semantic retrieval using a Qdrant vector store.","best_when":"You need agent-accessible, multi-domain OSINT-style intelligence enrichment and retrieval over many heterogeneous sources, and you can run/operate the MCP server yourself.","avoid_when":"You need strict compliance guarantees, strongly authenticated enterprise-grade access control, or you cannot tolerate uneven source reliability/latency across many third-party APIs.","last_evaluated":"2026-03-30T15:36:43.422159+00:00","has_mcp":true,"has_api":false,"auth_methods":[],"has_free_tier":false,"known_gotchas":["No documented auth; if you expose the MCP server remotely, you may need to add your own network controls.","Tool reliability depends on many upstream public APIs and RSS feeds; expect intermittent failures/timeouts.","Semantic/vector results quality depends on collector cadence and embedding/vector-store configuration.","Large number of tools (110+) increases the chance an agent selects an unsupported/edge-case tool unless constrained."],"error_quality":0.0}