Scientific Tools MCP Server
High-value scientific tools for AI agents — literature search (PubMed, arXiv, Semantic Scholar), chemical compound lookup (PubChem, ChEMBL), patent prior art search (USPTO, EPO), GPU spot prices across 4 providers, and real-time earth science data (USGS, NASA, OpenAQ). Per-call billing via API key. Keys issued instantly at https://mcp-site.com/keys/request
README
Scientific Tools MCP Server
An MCP-compatible server providing high-value scientific tools for AI agents. Earn per tool call. Zero upfront capital. Fully automated.
Tools
| Tool | Price/call | Description |
|---|---|---|
literature_search |
$0.020 | PubMed + arXiv + Semantic Scholar |
compound_lookup |
$0.010 | PubChem + ChEMBL molecular properties |
gpu_spot_prices |
$0.005 | Live GPU spot prices across 4 providers |
patent_prior_art_search |
$0.050 | USPTO + EPO patent databases |
scientific_data |
$0.003 | USGS earthquakes, NASA, OpenAQ air quality |
analytics |
$0.001 | Usage stats and revenue reporting |
Quick start
# 1. Install dependencies
pip install fastapi uvicorn stripe pydantic aiohttp
# 2. Set environment variables
cp .env.example .env
# Edit .env with your keys
# 3. Run demo (no keys needed)
python main.py
# 4. Start production server
python main.py --serve
Environment variables
# Required for billing
STRIPE_SECRET_KEY=sk_live_...
STRIPE_WEBHOOK_SECRET=whsec_...
# Stripe metered price IDs (create in Stripe dashboard)
STRIPE_PRICE_LITERATURE=price_...
STRIPE_PRICE_COMPOUND=price_...
STRIPE_PRICE_GPU=price_...
STRIPE_PRICE_PATENT=price_...
STRIPE_PRICE_SCIDATA=price_...
STRIPE_PRICE_ANALYTICS=price_...
# Optional — improves data access limits
NOAA_API_TOKEN=... # https://www.ncdc.noaa.gov/cdo-web/token
NASA_API_KEY=... # https://api.nasa.gov/
Deployment (VPS)
# On a fresh Ubuntu 22.04 VPS ($6/mo Hetzner)
# Install Python
sudo apt update && sudo apt install python3-pip python3-venv nginx certbot -y
# Clone and install
git clone https://github.com/yourname/mcp-scientific-tools
cd mcp-scientific-tools
python3 -m venv venv && source venv/bin/activate
pip install -r requirements.txt
# Run as service
sudo nano /etc/systemd/system/mcp-server.service
# [Unit]
# Description=MCP Scientific Tools Server
# [Service]
# WorkingDirectory=/home/ubuntu/mcp-scientific-tools
# ExecStart=/home/ubuntu/mcp-scientific-tools/venv/bin/python main.py --serve
# Restart=always
# [Install]
# WantedBy=multi-user.target
sudo systemctl enable mcp-server && sudo systemctl start mcp-server
# SSL via nginx + certbot
sudo certbot --nginx -d your-domain.com
Directory submissions (one-time, ~30 minutes)
-
mcp.so — https://mcp.so/submit Submit: server name, endpoint URL, tool descriptions
-
Glama.ai — https://glama.ai/mcp/servers/submit Submit: GitHub repo URL
-
Anthropic MCP directory — https://github.com/modelcontextprotocol/servers Open a PR adding your server to the README
-
GitHub topic — Add
mcp-servertag to your repo Automatically discoverable by agents scanning GitHub
Stripe setup
- Create a Stripe account at stripe.com
- Create a product: "Scientific Tools MCP"
- For each tool, create a metered price:
- Billing: recurring, monthly
- Metered usage: sum of usage during period
- Price per unit: tool's per-call rate
- Copy each price ID (price_...) to your .env
Agent integration example
import anthropic
client = anthropic.Anthropic()
# Add your MCP server
response = client.beta.messages.create(
model="claude-opus-4-6",
max_tokens=1024,
tools=[{
"type": "custom",
"name": "scientific_tools",
"description": "Scientific research tools",
"mcp_server": {
"url": "https://your-domain.com/mcp",
"auth": {"type": "bearer", "token": "mcp-key-..."}
}
}],
messages=[{
"role": "user",
"content": "Search for recent papers on CRISPR cancer therapy"
}]
)
Legal disclaimer and terms of service
BY USING THIS SERVICE OR ANY API KEY ISSUED BY THIS SERVICE, YOU AGREE TO THESE TERMS, INCLUDING THE DISCLAIMER THAT THE SERVICE IS PROVIDED AS IS WITHOUT WARRANTY, DATA IS SOURCED FROM THIRD-PARTY APIS AND MAY BE INACCURATE, AND OPERATOR LIABILITY IS LIMITED TO FEES PAID IN THE PRECEDING 30 DAYS. IF YOU DO NOT AGREE, DO NOT USE THIS SERVICE OR ITS API KEYS.
1. No warranties
This service is provided "as is" and "as available" without warranty of any kind, express or implied, including but not limited to warranties of merchantability, fitness for a particular purpose, or accuracy. The service retrieves data from third-party public APIs (PubMed, PubChem, USPTO, USGS, NASA, OpenAQ, and others) over which the operator has no control. That data may be incomplete, inaccurate, outdated, or unavailable at any time without notice.
2. Not professional advice
Data and outputs provided by this service do not constitute and must not be relied upon as medical, clinical, legal, patent, financial, pharmaceutical, toxicological, or safety advice. Users requiring professional advice must consult qualified licensed professionals. The operator expressly disclaims responsibility for any decisions made based on outputs from this service.
Specifically:
- Literature search results are not a substitute for systematic review by qualified researchers
- Compound property data is not a substitute for laboratory analysis by qualified chemists
- Patent search results are not a substitute for freedom-to-operate analysis by a licensed patent attorney
- Earthquake and environmental data are not a substitute for official emergency management guidance
3. Limitation of liability
To the maximum extent permitted by applicable law, the operator's aggregate liability for any claims arising from or related to this service shall not exceed the total fees paid by the claimant in the 30 days preceding the claim. The operator shall not be liable for any indirect, incidental, consequential, or punitive damages of any kind.
4. Data sources
This service retrieves data from: PubMed/NCBI, arXiv, Semantic Scholar, PubChem, ChEMBL, USPTO, EPO/Espacenet, USGS, NASA, and OpenAQ. Each source is subject to its own terms and accuracy limitations. The operator makes no representation regarding the accuracy or completeness of data from any of these sources.
5. Acceptable use
You agree not to use this service to make clinical, medical, or safety-critical decisions without independent professional verification, to circumvent rate limits or billing mechanisms, or to engage in any activity that violates applicable laws or regulations.
6. Service availability
The operator does not guarantee any specific level of uptime, availability, or response time. The service may be interrupted, modified, suspended, or discontinued at any time with or without notice.
7. Privacy
Query parameters are processed in memory solely to fulfil each request and are not stored beyond the duration of each individual request. Usage metadata (tool name, timestamp, call volume) is retained for billing and analytics. No personally identifiable information is knowingly collected or transmitted to third parties beyond what is required for payment processing (Stripe).
8. Governing law
These terms are governed by the laws of the United States and the state in which the operator is domiciled, without regard to conflict of law provisions.
Full terms of service: https://yourdomain.com/terms
Complete legal text: see DISCLAIMER.py in this repository
Revenue projection
At 5 tools × 1,000 calls/day × $0.01 avg price = $50/day = $1,500/month
Scale levers:
- More tools → more call volume
- Better tools → higher per-call price
- More agent integrations → more calls automatically
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.