Osdr MCP Server

This project demonstrates how to integrate MCP agents with NASA's Open Science Data Repository (OSDR) tools to achieve the fetching, analysis, and summary generation of biological data.

Research and data Developer tools #Data Agents #Biological Analysis #NASA Integration #Intelligent Summarization .Python

rating : 2 points

downloads : 4.1K

update time : 2025-04-29

Open Site

What is OSDR MCP Server?

A custom Model Context Protocol (MCP) server that provides specialized tools for interacting with NASA's Open Science Data Repository (OSDR). It enables AI agents to access and process biological research data through standardized interfaces.

How does it work?

The system uses AI agents that communicate through MCP to perform data operations. Agents can chain multiple tools together to complete complex workflows like fetching studies, analyzing RNA data, and generating reports.

Use Cases

Ideal for researchers and data scientists who need to: - Quickly explore NASA's biological datasets - Automate data analysis pipelines - Generate summaries of scientific findings

Key Features

OSDR Metadata Tools

Specialized tools for fetching and filtering NASA OSDR study metadata including by organism type

Visualization Tools

Built-in RNA data analysis and visualization capabilities

Multi-Agent Workflows

Pre-configured agent teams that collaborate on data tasks

Advantages

Simplifies access to NASA's biological research data

Pre-built analysis pipelines save development time

Flexible architecture supports custom tool integration

Limitations

Requires basic Python knowledge for customization

Currently focused on biological datasets

Limited to OSDR's available data

Getting Started

Install dependencies

Ensure you have Python installed, then install required packages

Run the MCP server

Launch the core service that provides OSDR access tools

Execute an agent

Run one of the example agents to see the system in action

Example Workflows

Quick Study Summary

Agent fetches study metadata and generates a tweet-length summary

RNA Data Analysis

Multi-agent workflow that downloads and visualizes RNA sequencing data

Frequently Asked Questions

What's the difference between the example agents?

Can I add my own analysis tools?

What LLMs are supported?

Additional Resources

Model Context Protocol Documentation

Official MCP specification and documentation

NASA OSDR API

Direct access to NASA's biological datasets

Example Repository

Source code for these examples

🚀 OSDR MCP Server and Agents

This repository demonstrates how to integrate MCP agents into custom tools for interacting with NASA's Open Science Data Repository (OSDR). It includes example agents for analyzing and summarizing biological data using the Model Context Protocol (MCP) workflow.

📦 Included Content

`osdr_mcp/`

A custom MCP server that provides tools for interacting with OSDR data:

osdr_fetch_metadata: Fetch metadata for a given OSDR dataset.
osdr_find_by_organism: Filter studies by organism.
Other tools (e.g., RNA analysis) are located in osdr_viz_tools.

`first_example/`

A simple agent that uses two official MCP servers:

mcp-server-fetch (headless browser)
mcp-server-filesystem

This agent collects information and generates a tweet-sized summary.

`second_example/`

Similar to first_example, but connects to the custom MCP server defined in osdr_mcp/main_simple.py. Demonstrates how to integrate domain-specific tools, such as the OSDR metadata retriever.

`third_example/`

A complete multi-agent workflow:

Fetch Agent - Scrape OSD study metadata.
Quantitative Analysis Agent - Download RNA count data and create bar charts.
Summary Writer Agent - Generate a Markdown report containing a summary of the analysis.

Uses custom MCP servers:

osdr_data_fetch
osdr_viz_tools

⚙️ Configuration

The cp_agent.config.yaml file controls:

LLM backends (e.g., Ollama, OpenAI)
MCP server connections
Tool availability
System prompts / metadata

🚀 Quick Start

Clone this repository.
Install dependencies:
```
pip install -r requirements.txt
```
Run an example:
```
python first_example/main.py
```

Or start the MCP server directly:

python osdr_mcp/main_simple.py

🧩 Integration Instructions

This architecture is designed to provide flexibility. You can switch between document Q&A, RAG search, or custom analysis tools as needed. For user interface integration, it's ideal to use mode switches or UI toggles. Milvus-based RAG support is under development.