Multimodal Information Synthesis for AI-Assisted Agricultural Field Assessment Using MCP
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
Multimodal Information Synthesis for AI-Assisted Agricultural Field Assessment Using Model Context Protocol
Author
Stefan Stiller
Leibniz-Centre for Agricultural Landscape Research (ZALF) e.V.
Email: stefan [dot] stiller [at] zalf [dot] de, stillsen [at] gmail [dot] com
ORCID: 0009-0004-7468-1678
Related MCP server: agribrain
Description
This repository accompanies the study "Multimodal Information Synthesis for AI-Assisted Agricultural Field Assessment Using Model Context Protocol".
It provides software for structured multimodal information access for large language models (LLMs) via the Model Context Protocol (MCP):
MCP resource servers that expose agricultural information modalities to LLM clients:
yield prediction from RGB imagery
scientific PDF literature (topic-based access)
soil CSV data (lab results and moisture)
Evaluation plotting code that regenerates the manuscript figures comparing:
structured MCP access vs unstructured availability of the same sources
increasing modality dose under MCP (single resources → pairs → all resources)
Related work on the underlying yield model is available at:
Self-Supervised Learning for Crop Classification and Yield Prediction
Repository contents
.
├── MCP_yield_server.py # MCP server: image-based yield prediction
├── MCP_pdf_resource_server.py # MCP server: topic-based PDF literature access
├── MCP_csv_resource_server.py # MCP server: soil CSV resources
├── llm_judge-two-question-figures.py # Manuscript Fig. 1 and Fig. 2
├── requirements.txt
├── CITATION.cff
├── LICENSE # GNU GPLv3
└── README.mdUsage
MCP servers
Start each server from an MCP-compatible client configuration (e.g. Claude Desktop / Cursor), pointing to the corresponding Python file:
MCP_yield_server.pyMCP_pdf_resource_server.pyMCP_csv_resource_server.py
Manuscript figures
python llm_judge-two-question-figures.pyThis writes Figure 1 (structured vs unstructured access) and Figure 2 (MCP modality ladder), plus caption text files, to Evaluation_Reports/.
Notes for reuse
Judge model names in the figures are remapped for display; scores come from the evaluation CSV referenced in the script.
Edit
OUTPUT_DIR/CSV_NAMEin the figure script if your layout differs.
Citation
If you use this code, please cite it as indicated in CITATION.cff (and the associated paper when available).
License
This project is licensed under the GNU GPLv3 License — see the LICENSE file for details.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
Access Pollinations models and API capabilities through agent tools.
Connect AI clients to biomedical data and tools.
Pay-per-call agent tools: PDF, OCR, translation, news search, prediction markets, web, chain reads.
Read satellite, NDVI, weather and soil data; create and delete fields in Agro API.
Related MCP Servers
AlicenseNot gradedqualityFmaintenanceEnables AI assistants to analyze farm locations, retrieve soil and weather data, access commodity market prices, and chat with an agricultural advisor for farming recommendations.MIT- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to access free, open agronomic data for field briefings, spray windows, water balance, pest pressure, and more, using only public data sources without API keys.1MIT
- FlicenseNot gradedqualityCmaintenanceEnables LLMs to search, preview, and query data assets from ICRISAT's federation-first data hub, supporting CSV, Excel, and Word files with SQL access via DuckDB.-
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to seal, open, verify, and ingest verifiable agri-traceability data by description, with credentials and private keys kept out of the model's flow.23 npmMIT