Privacy-preserving AI for global agriculture and farming.
Sovereign Mohawk Agriculture Global enables policy-gated federated learning across farms, co-ops, and regional operators while preserving full farmer data sovereignty. It extends the Sovereign Mohawk ecosystem from foundation and scale to food security and climate resilience.
Sovereign Mohawk Agriculture Global: every farm keeps full data sovereignty while contributing to planetary-scale AI for sustainable yields, climate resilience, pest and disease detection, soil health, and precision resource management.
Core repositories:
- Sovereign-Mohawk-Proto
- Sovereign_Map_Federated_Learning
- Sovereign_Mohawk_Oncology_Global
- Sovereign_Mohawk_Agriculture_Global
- Sovereign_Mohawk_Climate_Global
- Sovereign_Mohawk_SupplyChain_Global
Canonical architecture and shared execution plans:
- docs/ECOSYSTEM_ARCHITECTURE.md
- docs/UNIFIED_PHASE2_MILESTONES.md
- docs/GOVERNANCE_PR_ONLY.md
- docs/ACTION_PINNING_POLICY.md
- docs/PRODUCTION_DATA_PLANE_PLAN.md
- Federated learning across millions of edge nodes (tractors, drones, IoT sensors, weather stations, and satellite-derived field data) without centralizing sensitive farm data.
- Preserve farmer ownership of raw data (yields, soil scans, machinery telemetry, and proprietary farm practices).
- Enforce policy-gated model updates with Mohawk Protocol controls: BFT, zk-SNARK verification, TPM attestation, and PQC-ready transport.
- Deliver an interactive geospatial dashboard with farm-region layering and operational telemetry.
- Provide compliance automation for cross-border agricultural data governance.
This repository is the agriculture flagship in the progression:
Proto (foundation) -> Map (scalable FL engine) -> Oncology (healthcare) -> Agriculture (food security and climate).
- Flower security wrapper with signed-update policy enforcement.
- Mohawk streaming aggregation profile for high-scale rounds.
- Poisoning detection, differential privacy controls, and secure aggregation guardrails.
- Continent -> country -> cooperative region drill-down simulation.
- Crop-intelligence and yield-convergence visualizations.
- Soil health and climate risk overlays.
- FL pipeline, node health, tokenomics telemetry, and compliance scorecards.
- Agriculture control mapping (EU CAP-aligned handling, USDA-oriented privacy constraints, GDPR-style controls for farm-linked data).
- DPIA-style templates for farm data processing flows.
- Consent and regional policy engine patterns.
- Safe prompt and audit workflow templates for agronomic insights.
- Governance checklist before model-assisted recommendations.
- Threat-model lens tuned for ag-sector adversaries and supply-chain manipulation.
- MHC utility incentive model for honest FL contributions.
- Contribution-aware telemetry and operator scorecards.
index.html: dashboard entry point.app.js: dashboard logic, geospatial simulation, telemetry views.styles.css: visual theme and responsive layout.manifesto.html: agriculture edition manifesto.flower_security_wrapper/: policy wrapper templates.docs/: beta release, compliance evidence, manifesto, roadmap.scripts/: artifact capture and utility scripts.examples/: example farm sensor FL workflows..github/workflows/: static site deployment workflow.
Serve the dashboard locally:
cd Sovereign_Mohawk_Agriculture_Global
python3 -m http.server 8080Then open:
docker compose -f docker-compose.mock.yml upDashboard:
Mock API:
- Copy policy and streaming aggregation adapters from Map and Oncology wrappers.
- Wire live metrics endpoints to Prometheus/Grafana sources.
- Connect compliance controls to production policy definitions.
- Add CI quality gates (lint/test/security/performance) mirroring ecosystem standards.
The dashboard now supports live endpoint integration with simulation fallback.
Default endpoint targets:
/api/agri/stats/api/agri/regions/api/agri/pipeline/api/agri/risk/api/agri/health
You can override endpoints by setting window.AGRI_DASHBOARD_CONFIG before loading app.js.
Example:
<script>
window.AGRI_DASHBOARD_CONFIG = {
statsUrl: "https://your-gateway.example.com/agri/stats",
regionsUrl: "https://your-gateway.example.com/agri/regions",
pipelineUrl: "https://your-gateway.example.com/agri/pipeline",
riskUrl: "https://your-gateway.example.com/agri/risk",
healthUrl: "https://your-gateway.example.com/agri/health",
timeoutMs: 5000
};
</script>See docs/PHASE2_INTEGRATION.md for payload contract examples.
Completed in this sprint:
- Mock live API (
mock_api/server.py) with endpoint fixtures. - Contract validation script (
scripts/validate_dashboard_contract.py). - Contract test (
tests/test_contract_validator.py). - CI workflow (
.github/workflows/ci.yml) with contract and endpoint smoke checks. - Security workflows (
.github/workflows/codeql.yml,.github/workflows/secret-scan.yml). - Dependency update automation (
.github/dependabot.yml). - PR governance files (
.github/pull_request_template.md,.github/CODEOWNERS). - Threat baseline (
docs/THREAT_MODEL_BASELINE.md).
Run local validation:
python3 scripts/validate_dashboard_contract.py
python3 -m unittest -q tests/test_contract_validator.pyBefore cutting the first release tag:
- Ensure Pages deploy from
mainis green. - Ensure CI, CodeQL, and Secret Scan are green.
- Ensure contract validator passes with production-shaped payloads.
- Ensure branch protection required checks match active workflows.
Bootstrap scaffold complete. Domain adapters and production connectors are planned in the beta hardening cycle documented in docs/ROADMAP.md.