Marc Shade 8d60a9a220 feat: add core analysis engine — focal points, temporal anomalies, CII v2
Phase 5 of world-intel-mcp: intelligence synthesis layer that makes raw
data actionable.

New modules:
- config/countries.py: 22 tier-1 countries with event multipliers, 22
  intel hotspots, 9 strategic waterways, 6 conflict zones
- analysis/focal_points.py: detect convergence of signals on entities
  using signal count, type diversity, and recency weighting
- analysis/temporal.py: Welford's algorithm + SQLite for streaming
  anomaly detection with seasonal baselines (weekday+month)

Upgraded modules:
- analysis/instability.py: CII v2 with 4 weighted domains (unrest 0.25,
  conflict 0.30, security 0.20, information 0.25), UCDP floors, event
  multipliers. Full v1 backward compat preserved.
- analysis/signals.py: v2 aggregator accepts 7 data sources (added
  outages, military, protests) with convergence scoring
- sources/intelligence.py: 3 new fetch functions (focal_points,
  signal_summary, temporal_anomalies), instability upgraded to CII v2

3 new MCP tools: intel_focal_points, intel_signal_summary,
intel_temporal_anomalies. Total: 39 tools.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-23 18:39:57 -05:00

world-intel-mcp

Real-time global intelligence MCP server for Phoenix AGI. Covers financial markets, economic indicators, earthquakes, wildfires, conflict, military flights, infrastructure health, and more.

Quick Start

pip install -e .
world-intel-mcp        # Run as MCP server
intel markets           # CLI: stock indices
intel earthquakes       # CLI: recent quakes
S
Description
100+ tool MCP server for real-time global intelligence — markets, FX, bonds, earnings, SEC filings, conflict, military, cyber, climate, news, company enrichment, and 30+ domains. Live Leaflet dashboard with 20 map layers, SSE streaming, and AI situation briefs.
Readme MIT 1.4 MiB
Languages
Python 85.6%
HTML 14.1%
Shell 0.3%