Global Risk · Food Security · Natural Resource Management
01
How to Build This
This plan turns the Master Architecture (v1.7) and SIG-NAL's GeoAI Strategy into a single build order: one shared data-and-capability layer, one gateway, and three platform domains that pull from it rather than duplicate it. Global Risk, Food Security, and NRM are scoping and funding lenses, not three codebases — every capability below is built once, wrapped once, and registered once, regardless of which platform's budget paid for it.
Principle 01
One system, three domains
A government analyst, a hub scientist, and a development-partner engineer all pull from the same shared layer to build whatever solution their problem calls for.
Principle 02
Wrap once, reuse everywhere
Every capability — internal or partner-owned — becomes an MCP server once, then any authorized agent across all three domains can call it via Agent2Agent.
Principle 03
Vendor-agnostic by policy
Agents are defined by their MCP-wrapped tools and encoded domain/workflow expertise — callable by Claude, GPT-class models, Gemini, or an open-weight model on SOCRATES.
Principle 04
Build toward SERVIR-AI
github.com/SERVIR-AI is the confirmed consolidation org — the master gateway everything migrates into, not one option among several (Decision #14).
The shared system, top to bottom
Fig. 1 — One shared stack, three domains on top. Section numbers point to where each layer is specified below.
What this plan is, precisely. A build order for the shared gateway and capability layer, the per-platform agents that sit on top of it, and the sequencing that gets from today's scattered hub repos to the SERVIR-AI-consolidated system — grounded entirely in Master Architecture v1.7 and the GeoAI Strategy document, with nothing invented that wasn't already decided or already researched this project.