Skip to main content
Glama

MCP Quoting System

by r-long
TEST-CASES.md7 kB
# Test RFQ Examples This file contains sample RFQ requests you can use to test the system. ## How to Use These Tests ### Option 1: Using curl (Git Bash or PowerShell) ```bash curl -X POST http://localhost:3789/mcp/invoke/evaluateRfpAndDraftQuote \ -H "Content-Type: application/json" \ -d @test-rfq.json ``` ### Option 2: Using PowerShell ```powershell $body = Get-Content test-rfq.json -Raw Invoke-RestMethod -Uri http://localhost:3789/mcp/invoke/evaluateRfpAndDraftQuote ` -Method Post -Body $body -ContentType "application/json" ``` ### Option 3: Edit test-rfq.json and use quick-test.bat Just modify the test-rfq.json file and run the quick-test batch file. --- ## Test Case 1: High Similarity Match (Expected: 85%+ match) Should match Q-1001 in the database (6061-T6 aluminum, CNC, anodize) ```json { "rfp": { "rawText": "We need 200 pieces of 6061-T6 aluminum brackets, CNC machined with anodized finish. Tolerance: +/-0.005 inches. Delivery needed by February 28, 2025.", "qty": 200, "contactEmail": "buyer@acme.com", "customerName": "Acme Manufacturing" } } ``` **Expected Results:** - High confidence match - ~$12.50/unit cost estimate - ~21 day lead time - Match with Quote Q-1001 --- ## Test Case 2: Medium Similarity (Expected: 70-85% match) Similar material, different processes ```json { "rfp": { "rawText": "Need 75 units of 6061 aluminum plates, CNC milled and deburred. Standard tolerances +/-0.010. No special finish required.", "qty": 75, "contactEmail": "engineer@precision.com", "customerName": "Precision Parts Co" } } ``` **Expected Results:** - Medium confidence - Match with Q-1234 (similar aluminum CNC work) - ~$18-20/unit cost estimate --- ## Test Case 3: Stainless Steel with Welding (Expected: Match Q-0832) ```json { "rfp": { "rawText": "Request quote for 30 stainless steel 304 assemblies. Laser cut, TIG welded, and passivated. Tight tolerances +/-0.010. Marine application.", "qty": 30, "contactEmail": "orders@marine.com", "customerName": "Marine Systems Inc" } } ``` **Expected Results:** - Medium-high confidence - Match with Q-0832 (304 SS, laser, weld, passivate) - Higher cost due to welding (~$45-50/unit) - Longer lead time (~35 days) --- ## Test Case 4: High Volume Aluminum (Expected: Match Q-0955) ```json { "rfp": { "rawText": "Quote needed for 500 aluminum 6061 components, CNC machined with powder coat finish. Color: RAL 7035. Standard tolerances acceptable.", "qty": 500, "contactEmail": "purchasing@buildtech.com", "customerName": "BuildTech Solutions" } } ``` **Expected Results:** - High confidence match - Match with Q-0955 (high volume aluminum + powder coat) - Lower unit cost due to volume (~$8-9/unit) - 28 day lead time --- ## Test Case 5: Low Similarity - New Material (Expected: <70% match) ```json { "rfp": { "rawText": "We need 100 titanium Grade 5 brackets, CNC machined with precision grinding. Aerospace tolerances +/-0.002. AS9100 certification required.", "qty": 100, "contactEmail": "supply@aerospace.com", "customerName": "Aerospace Components LLC" } } ``` **Expected Results:** - Low confidence (new material: titanium) - No strong historical matches - Higher cost estimate due to material - Requires engineer review before quoting --- ## Test Case 6: Minimal Information (Expected: Low confidence) ```json { "rfp": { "rawText": "Need some metal parts. About 50 pieces. Let me know cost and time.", "qty": 50, "contactEmail": "info@startup.com" } } ``` **Expected Results:** - Low confidence (insufficient detail) - Generic cost estimate - System will request more information - Requires engineer clarification --- ## Test Case 7: Marine Grade Stainless (Expected: Match Q-1567) ```json { "rfp": { "rawText": "RFQ for 25 units 316 stainless steel fittings. CNC machined, passivated, and polished to mirror finish. Marine grade corrosion resistance required. Tolerances +/-0.005.", "qty": 25, "contactEmail": "marine@offshore.com", "customerName": "Offshore Marine Equipment" } } ``` **Expected Results:** - High confidence match - Match with Q-1567 (316 SS, passivate, polish) - Premium pricing (~$52/unit) due to 316 grade and polishing - 21 day lead time --- ## How to Interpret Results ### Confidence Levels - **High (≥85%)**: We've done very similar work before. Estimate is reliable. - **Medium (70-85%)**: Similar to past work but with differences. Review recommended. - **Low (<70%)**: New type of work or insufficient information. Engineer review required. ### Match Score Components The system evaluates similarity across: 1. **Material** (35% weight): Exact, family, or partial material match 2. **Processes** (30% weight): Overlap in manufacturing processes 3. **Quantity** (20% weight): Same or adjacent quantity ranges 4. **Tolerances** (10% weight): Precision requirement alignment 5. **Finish** (5% weight): Surface treatment matching ### Cost Estimate Includes: - Material cost - Processing cost (CNC time, etc.) - Labor cost - Tooling amortization - Overhead (15%) - Margin (20%) - Contingency (10% for low confidence quotes) ### Lead Time Factors Based on: - Material procurement time - Setup and tooling time - Run time (cycle time × quantity) - QA and inspection - Finishing processes - Shipping time --- ## Creating Your Own Test Cases Edit test-rfq.json with your own data. Include: **Required:** - `rawText`: Description of what you need - `qty`: Quantity needed **Optional but Recommended:** - `contactEmail`: Email address - `customerName`: Customer name - `dueDate`: Required delivery date (ISO format: YYYY-MM-DD) - `partNumber`: Part number if available - `material`: If known - `processes`: Array of process names **Good rawText descriptions include:** - Material specification (6061-T6, 304 SS, etc.) - Manufacturing processes (CNC, laser, weld, etc.) - Tolerances (+/-0.005, standard, precision, etc.) - Surface finish (anodize, powder coat, polish, etc.) - Special requirements (aerospace, marine, etc.) --- ## Troubleshooting Test Results ### "No similar quotes found" - Check that data/quotes.json has content - Try with materials in the sample database (6061, 304 SS, aluminum) - Make sure the server is running ### "Low confidence" when expecting high - Add more detail to rawText (material grade, processes) - Check material name spelling (6061-T6, not 6061T6) - Verify processes match database (CNC, anodize, etc.) ### Cost estimates seem off - Review config.ts for material prices - Check machine hour rate ($80/hr default) - Verify overhead and margin percentages - Consider quantity range effects ### Server errors - Check that server is running (http://localhost:3789/health) - Verify JSON syntax is valid - Look at server console for error messages - Try the quick-test.bat for automated testing --- **Happy Testing!** 🧪 For more information, see the main README.md or QUICKSTART.md files.

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/r-long/mcp-quoting-system'

If you have feedback or need assistance with the MCP directory API, please join our Discord server