Catalysis Hub MCP Server
A Model Context Protocol (MCP) server interface to Catalysis Hub's GraphQL API, enabling programmatic access to catalysis research data through flexible GraphQL queries.
Key Features
- Direct GraphQL Access: Execute any valid GraphQL query against Catalysis Hub's API
- Comprehensive Data Access:
- Catalytic reactions (equations, conditions, catalysts)
- Material systems (structures, properties, descriptors)
- Research publications (titles, DOIs, authors)
- Surface reaction data (adsorption energies, binding sites)
- MCP Standard Compliance: Implements the Model Context Protocol for AI-agent interoperability
- Flexible Query Support: Execute complex queries with variables parameterization
- Error Handling: Robust error reporting for API connectivity and query execution
License and Citation
This project is available under the MIT License with an Academic Citation Requirement. This means you can freely use, modify, and distribute the code, but any academic or scientific publication that uses this software must provide appropriate attribution.
For academic/research use:
If you use this software in a research project that leads to a publication, presentation, or report, you must cite this work according to the format provided in CITATION.md.
For commercial/non-academic use:
Commercial and non-academic use follows the standard MIT License terms without the citation requirement.
By using this software, you agree to these terms. See LICENSE.md for the complete license text.
Implementation Details
- Server Configuration (matches
claude_desktop_config.json):{ "command": "/Users/quentincody/.env/bin/python3", "args": ["/Users/quentincody/catalysishub-mcp-server/catalysishub_mcp_server.py"], "options": { "cwd": "/Users/quentincody/catalysishub-mcp-server" } } - Core Dependency:
httpxfor asynchronous HTTP requests - Transport: Standard input/output communication following MCP specifications
Setup & Installation
-
Clone the repository:
git clone <repository_url> cd catalysishub-mcp-server -
Install dependencies:
pip install -r requirements.txt -
Verify installation:
python3 catalysishub_mcp_server.py --version # Should output: catalysishub-mcp-server 0.1.0
Usage Examples
Basic Query Execution
from mcp.client import MCPClient
async with MCPClient("catalysishub") as hub:
result = await hub.catalysishub_graphql(
query="""{ reactions(first: 5) {
edges {
node {
id
Equation
Temperature
}
}
}
}"""
)
print(json.loads(result))
Parameterized Query with Variables
variables = {
"materialId": "mp-1234",
"firstResults": 5
}
response = await hub.catalysishub_graphql(
query="""query GetMaterial($materialId: String!, $firstResults: Int!) {
systems(uniqueId: $materialId) {
edges {
node {
energy
Cifdata
relatedReactions(first: $firstResults) {
edges {
node {
id
Equation
}
}
}
}
}
}
}""",
variables=variables
)
Query Optimization Tips
-
Use GraphQL Fragments:
fragment ReactionDetails on Reaction { id Equation ActivationEnergy Catalyst { formula surface } } query { reactions(first: 10) { edges { node { ...ReactionDetails } } } } -
Batch Related Queries:
query BatchQuery { reactions: reactions(first: 5) { edges { node { id Equation } } } materials: systems(first: 5) { edges { node { formula energy } } } }
Response Structure
Successful responses follow this structure:
{
"data": { /* Query results */ },
"extensions": {
"responseMetadata": {
"requestDuration": 145,
"apiVersion": "2024-06"
}
}
}
Error responses include detailed diagnostics:
{
"errors": [{
"message": "Cannot query field 'invalidField' on type 'Reaction'",
"locations": [{"line": 5, "column": 21}],
"path": ["query", "reactions", "edges", "node", "invalidField"]
}]
}
Troubleshooting
Common Issues:
HTTP Request Error: Verify network connectivity toapi.catalysis-hub.orgJSON Decode Error: Check query syntax using Catalysis Hub's GraphQL PlaygroundTimeout Errors: Addtimeoutparameter to complex queries
Acknowledgements
This project builds on the Model Context Protocol (MCP) framework and is designed to interface with the Catalysis Hub database, a comprehensive resource for catalysis research data.
Recommend MCP Servers 💡
mcp-zotero
A connector for Claude Desktop to work with collection and sources on your Zotero Cloud.
akr4/claude-code-mcp-docker
A Dockerized Claude Code MCP server designed for secure code execution, providing an isolated development environment for AI interactions.
Inflectra Spira Server
A Model Context Protocol (MCP) server enabling AI assistants to interact with the Inflectra Spira platform, providing a bridge between natural language interactions and the Spira REST API.
mcp-panther
Write detections, investigate alerts, and query logs—all from your favorite AI agents
apollographql/apollo-mcp-server
Apollo MCP Server enables AI assistants like Claude Desktop to interact with any GraphQL API by exposing GraphQL operations as AI-accessible tools, facilitating real-time data queries.
moodle-mcp
An MCP server implementation for interacting with Moodle LMS, providing access to Moodle's web service API functions.