warmthengine.com/mcp
MCP server exposing 16 programmatic tools for AI systems to query verified Coordination Intelligence on AI infrastructure events, connections, and actors across geopolitical blocs.
README
Warmth Engine Observatory (WEO)
AI Infrastructure Coordination Dynamics
Overview
The Warmth Engine Observatory tracks AI infrastructure coordination dynamics across geopolitical blocs through verified events and mapped Coordination Connections.
Scope: WEO tracks systemic AI infrastructure coordination — observable coordination dynamics between major power centres (currently nation-states and geopolitical blocs). This differs from operational AI infrastructure management, which addresses how organisations run their own AI systems.
Live Platform: www.warmthengine.com
Platform Components
- Event Database — Verified AI infrastructure coordination events classified by tier (Cross-Bloc, Between-Bloc, Intra-Bloc) and domain (Energy/Infrastructure, Security, Technology, Trade), tagged by industry sector (11-category GICS-anchored taxonomy) and assessed for environmental intersections via Environmental Nexus Tags (5-type ENT taxonomy with Warmth Engine Ratio)
- Coordination Connections — Documented relationships between events revealing structural patterns across the database (Policy Cascade, Joint Issuance, Funding Flow, Regulatory Enablement, Response, Framework Family, Parallel Policy)
- Interactive Map — Geographic visualisation of global AI infrastructure deployment
- Interactive Atlas — Network, timeline, and connectedness analytical surface for exploring coordination topology and event relationships via Cytoscape.js
- Evidence-Based Verification — Every event verified through primary and corroborating sources with gold standard evidentiary requirements
- Sovereign Capability Profiles — 7-dimension framework assessing 20 state actors' sovereign control across the AI infrastructure stack, with four designation levels (PAA, AIK, ACS, Participant); includes qualification headlines, constraint summaries, sourced evidence, and severance analysis
- Profiles Page — Standalone
/profiles/page rendering the full SCP panel with dimension filtering, actor detail expansion, and Dimension Watch table - Dimension Watch — Forward-looking monitoring table identifying actors approaching a designation status change at the next assessment cycle, with trigger conditions and timeline estimates
- Infrastructure Threads — Factual cross-event links documenting where the same AI infrastructure asset (chip or accelerator, energy source, or fabrication facility) appears across multiple verified events, with four link types (fabricated_at, deploys, trains_on, powered_by); toggleable overlay on the Atlas and queryable via
get_threads - Guided Journeys — Operator-curated analytical pathways across three substrates on the Atlas: coordination (directional coordination lineage), hardware (infrastructure thread links), and all layers (fusion traversal spanning coordination, material, and capability layers)
- Coordination Topology — Parent-child coordination backbone mapping directional relationships (Policy Cascade, Funding Flow, Regulatory Enablement, Response) across the event corpus, powering the Atlas network view
- MCP Integration — Model Context Protocol server at
warmthengine.com/mcpexposing sixteen programmatic tools for AI systems to query verified Coordination Intelligence (search_events, get_event, get_connections, get_connection_by_id, get_actors, get_coverage_stats, get_methodology, get_topology, traverse_coordination, get_capability_signatures, get_capability_links, get_threads, get_event_stack, query_scp, get_blocs, get_contribution)
Architecture
The platform uses a decoupled architecture: HTML files serve as lightweight UI shells that fetch data dynamically from the WEO API (Cloudflare Worker + KV).
- Data source: Cloudflare KV via WEO API (
warmthengine.com/api/) - MCP endpoint:
warmthengine.com/mcp(Model Context Protocol for AI agent access) - Authentication: Token-based (
?token=weo_xxx) for API and MCP consumers; password header for website - Fallback: Static JSON file for API downtime resilience
- Hosting: GitHub Pages with custom domain
- Routes: Managed via
wrangler.toml(warmthengine.com/api/*andwarmthengine.com/mcp*)
Data updates flow through the API — no HTML changes required for new events or connections.
Methodology Documentation
The WEO methodology is documented across four publications, all available on Zenodo with DOI registration:
Core Documentation
| Publication | DOI |
|---|---|
| WEO Methodology Manual — Operational specification covering event verification, significance assessment, tier classification, domain classification, industry tagging, environmental nexus tagging, and Coordination Connection methodology | 10.5281/zenodo.18427565 |
| WEO Methodology Rationale — Companion document providing empirical derivation, theoretical grounding, and calibration justification for all methodology components | 10.5281/zenodo.18427582 |
Methodology Extensions
| Publication | DOI |
|---|---|
| Towards Coordination Science: A Framework for Measuring Geopolitical Coordination Dynamics Through Event Relationship Patterns — Coordination Connection metrics and theoretical foundations (Layer 2 framework) | 10.5281/zenodo.18427584 |
| Tier-Crossing Dynamics in Coordination Networks: A Methodology Extension for Cross-Level Connection Analysis — Analytical framework for examining Coordination Connections across different tiers | 10.5281/zenodo.18427586 |
Self-Hosted Documentation
The V1.5 Manual and Rationale are self-hosted as navigable HTML with granular section anchoring:
- Manual:
warmthengine.com/research/methodology/manual/(490 section anchors) - Rationale:
warmthengine.com/research/methodology/rationale/(193 section anchors) - Errata & Corrigenda:
warmthengine.com/research/methodology/errata/— V1.5 errata and corrigenda (expression corrections and substantive corrections to reported findings, with machine-readableerrata.json)
A semantic dictionary (data/weo-methodology-map.json) maps 185 WEO-specific terms to their methodology section anchors, serving both human tooltip navigation and the MCP get_methodology tool.
Repository Structure
warmth-engine-observatory/
├── index.html # Main platform (interactive map)
├── atlas.html # Interactive atlas (network/timeline/connectedness)
├── events.html # Events database
├── methodology.html # Methodology overview
├── blocs.html # Bloc analysis
├── research.html # Research publications
├── about.html # About WEO
├── support.html # Supporter access
├── legal.html # Legal information
├── 404.html # Custom error page
├── profiles/
│ └── index.html # Sovereign Capability Profiles page (generated)
├── scripts/
│ └── generate_profiles.py # Profiles page generator (actors JSON → HTML)
├── data/
│ ├── events-free.json # Static fallback (free-tier event data)
│ └── weo-methodology-map.json # 185 term-to-anchor semantic mappings
├── research/
│ └── methodology/
│ ├── errata/ # V1.5 Errata Register (HTML + errata.json)
│ ├── manual/ # Self-hosted V1.5 Manual (HTML + PDF)
│ ├── materials-register/ # AI-Critical Materials Register (HTML)
│ └── rationale/ # Self-hosted V1.5 Rationale (HTML + PDF)
├── sitemap.xml # XML sitemap for search engines
├── robots.txt # Crawler directives
├── og-image.png # Open Graph social preview image
├── favicon_48x48.png # Browser tab icon (48px)
├── favicon_32x32.png # Browser tab icon (32px)
├── favicon_16x16.png # Browser tab icon (16px)
├── CNAME # Custom domain configuration
├── LICENSE.md # Licence information
└── README.md # This file
Citation
If using this platform or its methodology in research, please cite:
Warmth Engine Observatory (WEO)
Warmth Engine, 2026
https://www.warmthengine.com
Licensed under CC BY 4.0
Contact
- Website: warmthengine.com
- Email: warmthengine@proton.me
Acknowledgements
This platform uses the following open-source libraries and typefaces:
- Leaflet (BSD-2-Clause) — Interactive map functionality
- Leaflet.markercluster (MIT) — Map marker clustering
- Geist (SIL Open Font Licence) — Primary typeface
- Geist Mono (SIL Open Font Licence) — Monospace typeface
- Cytoscape.js (MIT) — Network graph visualisation (Atlas)
Licence
Methodology documentation is licensed under Creative Commons Attribution 4.0 International.
The platform, event database, and data products are proprietary.
See LICENSE.md for full details.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。