Medical Research MCP Suite

# 🧪 Testing Guide for Medical Research MCP Suite ## **Quick Start Testing (5 minutes)** ### **1. Basic Functionality Test** ```bash cd /Users/eugenezhou/Code/medical-research-mcp-suite # Make test script executable and run it chmod +x test-mcp.sh ./test-mcp.sh ``` This will test: - ✅ MCP server starts correctly - ✅ Tools are listed properly - ✅ Clinical trials search works - ✅ PubMed search works - ✅ Cross-API analysis works --- ## **Manual Testing Commands** ### **Test 1: List All Available Tools** ```bash echo '{"method":"tools/list","params":{}}' | npm run dev ``` **Expected Output:** ```json { "tools": [ { "name": "ct_search_trials", "description": "Search ClinicalTrials.gov for studies with AI-enhanced results" }, { "name": "research_comprehensive_analysis", "description": "Comprehensive analysis combining clinical trials, literature, and safety data" } // ... more tools ] } ``` ### **Test 2: Search Clinical Trials** ```bash echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":"diabetes","intervention":"metformin","pageSize":5}}}' | npm run dev ``` **Expected Output:** - JSON response with clinical trials data - Study summaries and metadata - NCT IDs and study details ### **Test 3: Get Specific Study** ```bash echo '{"method":"tools/call","params":{"name":"ct_get_study","arguments":{"nctId":"NCT04373031"}}}' | npm run dev ``` **Expected Output:** - Detailed study information for that specific NCT ID - Enhanced analysis and key points ### **Test 4: PubMed Literature Search** ```bash echo '{"method":"tools/call","params":{"name":"pm_search_papers","arguments":{"query":"diabetes AND metformin","maxResults":5}}}' | npm run dev ``` **Expected Output:** - PubMed papers related to diabetes and metformin - Paper titles, authors, abstracts ### **Test 5: FDA Drug Search** ```bash echo '{"method":"tools/call","params":{"name":"fda_search_drugs","arguments":{"drugName":"metformin"}}}' | npm run dev ``` **Expected Output:** - FDA-approved drug information - Approval dates and details ### **Test 6: Cross-API Comprehensive Analysis** (🔥 **The Magic!**) ```bash echo '{"method":"tools/call","params":{"name":"research_comprehensive_analysis","arguments":{"drugName":"pembrolizumab","condition":"lung cancer","analysisDepth":"detailed"}}}' | npm run dev ``` **Expected Output:** - Combined analysis from all databases - Risk assessments and market insights - Executive summary and recommendations ### **Test 7: Drug Safety Profile** ```bash echo '{"method":"tools/call","params":{"name":"research_drug_safety_profile","arguments":{"drugName":"metformin","timeframe":"5years","includeTrials":true,"includeFDA":true}}}' | npm run dev ``` **Expected Output:** - Comprehensive safety analysis - Clinical trial safety data - FDA adverse event data - Risk assessments and recommendations --- ## **Automated Test Suite** ### **Run All Tests** ```bash npm test ``` ### **Run Specific Test Files** ```bash # Test clinical trials API npm test -- clinicalTrials.test.ts # Run tests with coverage npm run test:coverage # Run tests in watch mode npm test -- --watch ``` ### **Test Individual Components** ```bash # Test API clients directly npm test -- --testNamePattern="ClinicalTrialsClient" # Test validation npm test -- --testNamePattern="Validator" # Test caching npm test -- --testNamePattern="Cache" ``` --- ## **Integration Testing with Claude Desktop** ### **1. Add to Claude Desktop Config** Add this to your `claude_desktop_config.json`: ```json { "mcpServers": { "medical-research": { "command": "node", "args": ["/Users/eugenezhou/Code/medical-research-mcp-suite/dist/index.js"], "env": { "NODE_ENV": "development" } } } } ``` ### **2. Restart Claude Desktop** - Close Claude Desktop completely - Reopen it - Look for "medical-research" in available tools ### **3. Test in Claude Desktop** Ask Claude: > "Can you search for clinical trials about diabetes using metformin? Use the medical research tools." > "Analyze the drug safety profile for pembrolizumab using the comprehensive analysis tool." > "What clinical trials are currently recruiting for lung cancer treatments?" --- ## **Performance Testing** ### **Test Response Times** ```bash # Time how long searches take time echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":"cancer","pageSize":10}}}' | npm run dev ``` ### **Test Caching** ```bash # Run same query twice - second should be faster echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":"diabetes","pageSize":5}}}' | npm run dev # Run again immediately (should be cached) echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":"diabetes","pageSize":5}}}' | npm run dev ``` ### **Test Error Handling** ```bash # Test invalid NCT ID echo '{"method":"tools/call","params":{"name":"ct_get_study","arguments":{"nctId":"INVALID"}}}' | npm run dev # Test empty search echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":""}}}' | npm run dev ``` --- ## **Load Testing** ### **Test Multiple Concurrent Requests** ```bash # Run multiple searches in parallel for i in {1..5}; do echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":"diabetes","pageSize":3}}}' | npm run dev & done wait ``` --- ## **Debugging & Logs** ### **Check Logs** ```bash # View all logs tail -f logs/combined.log # View only errors tail -f logs/error.log # Watch logs in real-time during testing tail -f logs/combined.log & npm run dev ``` ### **Debug Mode** ```bash # Run with debug logging LOG_LEVEL=debug npm run dev ``` ### **Test with Verbose Output** ```bash # Add debug info to your test echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":"diabetes","pageSize":3}}}' | LOG_LEVEL=debug npm run dev ``` --- ## **Common Issues & Solutions** ### **❌ "Command not found" errors** ```bash # Make sure you're in the right directory cd /Users/eugenezhou/Code/medical-research-mcp-suite # Make sure dependencies are installed npm install # Make sure project is built npm run build ``` ### **❌ API timeout errors** ```bash # Check your internet connection curl -I https://clinicaltrials.gov # Test with smaller queries echo '{"method":"tools/call","params":{"name":"ct_search_trials","arguments":{"condition":"diabetes","pageSize":1}}}' | npm run dev ``` ### **❌ JSON parsing errors** - Make sure JSON is properly formatted - Use single quotes for the outer shell command - Use double quotes inside the JSON ### **❌ Rate limiting** - Wait a few minutes between large requests - Use smaller page sizes - Check if you need API keys --- ## **Testing Checklist** ### **Basic Functionality ✅** - [ ] MCP server starts without errors - [ ] Tools list returns all expected tools - [ ] Clinical trials search works - [ ] PubMed search works - [ ] FDA search works - [ ] Cross-API analysis works ### **Data Quality ✅** - [ ] Search results contain valid data - [ ] NCT IDs are properly formatted - [ ] Papers have titles and abstracts - [ ] Error messages are helpful ### **Performance ✅** - [ ] Responses come back within 10 seconds - [ ] Caching works (second request faster) - [ ] Memory usage stays reasonable - [ ] No memory leaks over time ### **Error Handling ✅** - [ ] Invalid inputs return error messages - [ ] Network failures are handled gracefully - [ ] Rate limits don't crash the server - [ ] Malformed JSON is handled properly ### **Claude Desktop Integration ✅** - [ ] Server appears in Claude's tool list - [ ] Tools can be called from Claude - [ ] Responses are properly formatted - [ ] Claude can use the results effectively --- ## **🎯 Quick Success Test** Run this single command to verify everything works: ```bash cd /Users/eugenezhou/Code/medical-research-mcp-suite && npm run build && echo '{"method":"tools/call","params":{"name":"research_comprehensive_analysis","arguments":{"drugName":"metformin","condition":"diabetes","analysisDepth":"basic"}}}' | npm run dev ``` If this returns a comprehensive analysis combining clinical trials, literature, and FDA data, **your MCP server is working perfectly!** 🎉 --- **Ready to test? Start with the quick test script and work your way through the manual tests!**