Plug EvidenceMode into your agent

EvidenceMode is an MCP that gives your AI access to verified sources instead of hallucinating. You send a question, we return peer-reviewed studies with citations, journals, years, and disclaimers.

Your assistant can't answer without evidence. That's the point.

Peer-Reviewed Sources

Access PubMed, Cochrane Library, and NIH Reporter with full citations

Verified Evidence

No more AI hallucinations. Every response backed by real research

Easy Integration

Simple MCP integration with clear documentation and examples

Get your API key

Sign in to your dashboard

Go to /dashboard and copy your API key (starts with evd_live_...)

Use in your requests

Include your API key as x-api-key header in every request

curl -X POST https://evidencemode.xyz/api/evidence/search \
  -H "Content-Type: application/json" \
  -H "x-api-key: evd_live_your_key_here" \
  -d '{
    "question": "Is Ozempic safe long-term?",
    "domain": "health"
  }'

Understanding the response

Response Structure

Each response includes peer-reviewed summaries, journal citations, publication years, and required disclaimers.

{
  "results": [
    {
      "summary": "Semaglutide (Ozempic) produced sustained weight loss and improved glycemic control in multi-year adult trials with type 2 diabetes. Reported risks include GI effects and gallbladder-related events.",
      "journal": "New England Journal of Medicine",
      "year": 2021,
      "peer_reviewed": true,
      "link": "https://pubmed.ncbi.nlm.nih.gov/12345678/"
    }
  ],
  "disclaimer": "For informational purposes. Not medical advice."
}

Important Implementation Note

Your agent should surface the summary to the user, show the journal + year as a citation, and include the disclaimer. Do not edit the disclaimer out.

MCP Integration

Setup Instructions

Get Your API Key

Configure Your AI Development Environment

Add EvidenceMode to your preferred AI development platform:

Direct API Integration (Recommended)

Use EvidenceMode directly in your applications:

// In your application code
const evidence = await fetch("https://evidencemode.xyz/api/evidence/search", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "x-api-key": "evd_live_your_key_here"
  },
  body: JSON.stringify({
    question: "Is intermittent fasting safe for diabetics?",
    domain: "health"
  })
}).then(r => r.json());

Custom MCP Server

Create your own MCP server that connects to EvidenceMode:

// Create a simple MCP server
import { Server } from '@modelcontextprotocol/sdk/server/index.js';
import { StdioServerTransport } from '@modelcontextprotocol/sdk/server/stdio.js';

const server = new Server(
  {
    name: 'evidencemode-mcp',
    version: '1.0.0',
  },
  {
    capabilities: {
      tools: {},
    },
  }
);

// Add EvidenceMode tool
server.setRequestHandler('tools/call', async (request) => {
  if (request.params.name === 'evidence_search') {
    // Call EvidenceMode API here
    const response = await fetch('https://evidencemode.xyz/api/evidence/search', {
      method: 'POST',
      headers: {
        'Content-Type': 'application/json',
        'x-api-key': process.env.EVIDENCEMODE_API_KEY
      },
      body: JSON.stringify(request.params.arguments)
    });
    return await response.json();
  }
});

Claude Desktop Integration

Configure Claude Desktop to use your custom MCP server:

// In claude_desktop_config.json
{
  "mcpServers": {
    "evidencemode": {
      "command": "node",
      "args": ["path/to/your/evidencemode-mcp-server.js"],
      "env": {
        "EVIDENCEMODE_API_KEY": "evd_live_your_key_here"
      }
    }
  }
}

Test the Integration

Ask your AI assistant a factual question like "Is intermittent fasting safe?" and it will automatically use EvidenceMode to provide evidence-based answers.

When to Use EvidenceMode in 2025

Health & Medical AI

• Drug safety and efficacy research
• Treatment protocol validation
• Medical AI model training data
• Clinical decision support systems

AI Development & Engineering

• LLM training data validation
• AI model safety verification
• Framework security assessments
• Production system reliability

Finance & Economics

• Market analysis and trends
• Economic policy research
• Financial modeling validation
• Investment strategy research

Engineering & Technology

• Technical specification validation
• Engineering best practices
• Algorithm and system design
• Performance optimization research

Modern AI Development Workflows

Cursor IDE Integration

Get evidence-based code suggestions and architectural decisions

Claude Desktop Enhancement

Access peer-reviewed sources during AI conversations

VS Code MCP Extension

Real-time evidence validation in your development environment

AI Agent Training

Enhance your custom AI agents with verified knowledge

Domain Examples & Interaction Patterns

Health Domain

Example Questions:

• "Is intermittent fasting safe for diabetics?"
• "What are the side effects of metformin?"
• "Is Ozempic safe for long-term use?"

curl -X POST https://evidencemode.xyz/api/evidence/search \
  -H "Content-Type: application/json" \
  -H "x-api-key: evd_live_your_key_here" \
  -d '{"question": "Is intermittent fasting safe for diabetics?", "domain": "health"}'

Finance Domain

Example Questions:

• "What are the effects of inflation on stock markets?"
• "How do interest rates affect cryptocurrency prices?"
• "What is the impact of quantitative easing on bond yields?"

curl -X POST https://evidencemode.xyz/api/evidence/search \
  -H "Content-Type: application/json" \
  -H "x-api-key: evd_live_your_key_here" \
  -d '{"question": "What are the effects of inflation on stock markets?", "domain": "finance"}'

Engineering Domain

Example Questions:

• "What are the best practices for machine learning deployment?"
• "How to optimize database performance for large datasets?"
• "What are the security considerations for microservices?"

curl -X POST https://evidencemode.xyz/api/evidence/search \
  -H "Content-Type: application/json" \
  -H "x-api-key: evd_live_your_key_here" \
  -d '{"question": "What are the best practices for machine learning deployment?", "domain": "engineering"}'

Development Domain

Example Questions:

• "What is the best React state management library?"
• "How to implement secure authentication in Node.js?"
• "What are the performance benefits of Next.js SSR?"

curl -X POST https://evidencemode.xyz/api/evidence/search \
  -H "Content-Type: application/json" \
  -H "x-api-key: evd_live_your_key_here" \
  -d '{"question": "What is the best React state management library?", "domain": "dev"}'

Data Sources & Response Format

Data Sources

Health: PubMed (peer-reviewed medical journals)

Dev: GitHub repositories & technical documentation

Finance: arXiv finance papers & research

Engineering: arXiv technical papers & research

Response Format

{
  "results": [
    {
      "summary": "Evidence-based summary of findings",
      "journal": "Journal or publication name", 
      "year": 2024,
      "peer_reviewed": true,
      "link": "https://direct-link-to-source"
    }
  ],
  "disclaimer": "Domain-specific disclaimer text"
}

Rate Limits & Usage

Free Plan

calls/month

150

Pro Plan

calls/month

∞

Enterprise

unlimited

Real-World Integration Examples

React/Next.js Application

// pages/api/evidence.ts
import { NextApiRequest, NextApiResponse } from 'next';

export default async function handler(req: NextApiRequest, res: NextApiResponse) {
  const { question, domain } = req.body;
  
  const response = await fetch('https://evidencemode.xyz/api/evidence/search', {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json',
      'x-api-key': process.env.EVIDENCEMODE_API_KEY!
    },
    body: JSON.stringify({ question, domain })
  });
  
  const evidence = await response.json();
  res.json(evidence);
}

// In your React component
const getEvidence = async (question: string) => {
  const response = await fetch('/api/evidence', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({ question, domain: 'health' })
  });
  return response.json();
};

Python Application

# Python integration example
import requests
import os

def get_evidence(question: str, domain: str = "health"):
    response = requests.post(
        "https://evidencemode.xyz/api/evidence/search",
        headers={
            "Content-Type": "application/json",
            "x-api-key": os.getenv("EVIDENCEMODE_API_KEY")
        },
        json={"question": question, "domain": domain}
    )
    return response.json()

# Usage
evidence = get_evidence("Is intermittent fasting safe for diabetics?")
print(evidence["results"][0]["summary"])

Ready to get started?

Join developers building evidence-based AI applications

Get your API key