@andrewlwn77/nocodb-mcp

# NocoDB MCP Server - Basic Usage Examples This document provides practical examples of using the NocoDB MCP server for common operational tasks. ## Example 1: Project Management System ### Create a Projects Table ```json { "tool": "create_table", "arguments": { "base_id": "p_abc123", "table_name": "projects", "columns": [ { "title": "Project Name", "uidt": "SingleLineText", "rqd": true }, { "title": "Description", "uidt": "LongText" }, { "title": "Status", "uidt": "SingleSelect", "dtxp": "'planning','active','completed','on_hold'" }, { "title": "Start Date", "uidt": "Date" }, { "title": "Due Date", "uidt": "Date" }, { "title": "Budget", "uidt": "Currency" }, { "title": "Team Lead", "uidt": "SingleLineText" }, { "title": "Priority", "uidt": "Number", "dt": "int" } ] } } ``` ### Insert a Project ```json { "tool": "insert_record", "arguments": { "base_id": "p_abc123", "table_name": "projects", "data": { "Project Name": "AI Agent Integration", "Description": "Integrate NocoDB with AI agent teams for data sharing", "Status": "active", "Start Date": "2024-01-15", "Due Date": "2024-03-30", "Budget": 50000, "Team Lead": "Alice Johnson", "Priority": 1 } } } ``` ### Query Active Projects ```json { "tool": "query", "arguments": { "base_id": "p_abc123", "table_name": "projects", "where": "(Status,eq,active)", "sort": ["Priority", "-Due Date"], "fields": ["Project Name", "Team Lead", "Due Date", "Budget"] } } ``` ## Example 2: Customer CRM ### Create Customers Table ```json { "tool": "create_table", "arguments": { "base_id": "p_abc123", "table_name": "customers", "columns": [ { "title": "Company Name", "uidt": "SingleLineText", "rqd": true }, { "title": "Contact Email", "uidt": "Email", "unique": true }, { "title": "Phone", "uidt": "PhoneNumber" }, { "title": "Industry", "uidt": "SingleSelect", "dtxp": "'tech','finance','healthcare','retail','other'" }, { "title": "Annual Revenue", "uidt": "Currency" }, { "title": "Account Status", "uidt": "SingleSelect", "dtxp": "'prospect','active','churned'" }, { "title": "Last Contact", "uidt": "DateTime" }, { "title": "Notes", "uidt": "LongText" } ] } } ``` ### Bulk Insert Customers ```json { "tool": "bulk_insert", "arguments": { "base_id": "p_abc123", "table_name": "customers", "records": [ { "Company Name": "TechCorp Inc", "Contact Email": "contact@techcorp.com", "Phone": "+1-555-0100", "Industry": "tech", "Annual Revenue": 5000000, "Account Status": "active", "Last Contact": "2024-01-10T14:30:00Z" }, { "Company Name": "HealthFirst", "Contact Email": "info@healthfirst.com", "Phone": "+1-555-0200", "Industry": "healthcare", "Annual Revenue": 3000000, "Account Status": "active", "Last Contact": "2024-01-08T10:00:00Z" }, { "Company Name": "RetailMax", "Contact Email": "sales@retailmax.com", "Phone": "+1-555-0300", "Industry": "retail", "Annual Revenue": 8000000, "Account Status": "prospect", "Last Contact": "2024-01-05T16:45:00Z" } ] } } ``` ### Get Total Revenue by Industry ```json { "tool": "group_by", "arguments": { "base_id": "p_abc123", "table_name": "customers", "column_name": "Industry", "where": "(Account Status,eq,active)" } } ``` ## Example 3: Task Management ### Create Tasks Table with Relations ```json { "tool": "create_table", "arguments": { "base_id": "p_abc123", "table_name": "tasks", "columns": [ { "title": "Task Title", "uidt": "SingleLineText", "rqd": true }, { "title": "Description", "uidt": "LongText" }, { "title": "Assigned To", "uidt": "SingleLineText" }, { "title": "Due Date", "uidt": "Date" }, { "title": "Priority", "uidt": "SingleSelect", "dtxp": "'low','medium','high','urgent'" }, { "title": "Status", "uidt": "SingleSelect", "dtxp": "'todo','in_progress','review','done'" }, { "title": "Estimated Hours", "uidt": "Number", "dt": "float" }, { "title": "Completed", "uidt": "Checkbox" }, { "title": "Attachments", "uidt": "Attachment" } ] } } ``` ### Search Tasks ```json { "tool": "search_records", "arguments": { "base_id": "p_abc123", "table_name": "tasks", "query": "integration", "where": "(Status,neq,done)", "sort": "-Priority" } } ``` ### Update Task Status ```json { "tool": "update_record", "arguments": { "base_id": "p_abc123", "table_name": "tasks", "record_id": "rec_12345", "data": { "Status": "in_progress", "Assigned To": "Bob Smith" } } } ``` ## Example 4: Financial Tracking ### Create Invoices Table ```json { "tool": "create_table", "arguments": { "base_id": "p_abc123", "table_name": "invoices", "columns": [ { "title": "Invoice Number", "uidt": "SingleLineText", "unique": true, "rqd": true }, { "title": "Customer", "uidt": "SingleLineText", "rqd": true }, { "title": "Amount", "uidt": "Currency", "rqd": true }, { "title": "Tax", "uidt": "Currency" }, { "title": "Total", "uidt": "Formula", "formula": "{Amount} + {Tax}" }, { "title": "Issue Date", "uidt": "Date", "rqd": true }, { "title": "Due Date", "uidt": "Date", "rqd": true }, { "title": "Status", "uidt": "SingleSelect", "dtxp": "'draft','sent','paid','overdue','cancelled'" }, { "title": "Payment Date", "uidt": "Date" } ] } } ``` ### Get Total Outstanding Amount ```json { "tool": "aggregate", "arguments": { "base_id": "p_abc123", "table_name": "invoices", "column_name": "Total", "function": "sum", "where": "(Status,eq,sent)~or(Status,eq,overdue)" } } ``` ### Find Overdue Invoices ```json { "tool": "query", "arguments": { "base_id": "p_abc123", "table_name": "invoices", "where": "(Status,eq,sent)~and(Due Date,lt,2024-01-15)", "sort": ["Due Date"], "fields": ["Invoice Number", "Customer", "Total", "Due Date"] } } ``` ## Example 5: Creating Views ### Create a View for High Priority Tasks ```json { "tool": "create_view", "arguments": { "table_id": "tbl_tasks123", "title": "High Priority Tasks", "type": 1 } } ``` ### Get Data from View ```json { "tool": "get_view_data", "arguments": { "base_id": "p_abc123", "table_name": "tasks", "view_id": "vw_highpri456", "limit": 20 } } ``` ## Tips for AI Teams Using NocoDB 1. **Standardize Field Names**: Use consistent naming across tables (e.g., always use "created_at", "updated_at") 2. **Use Views for Common Queries**: Instead of complex filters, create views for frequently accessed data 3. **Leverage Formulas**: Use formula fields for calculated values instead of computing in your application 4. **Implement Soft Deletes**: Add a "deleted" checkbox field instead of actually deleting records 5. **Track Changes**: Add "created_by" and "updated_by" fields to track which AI agent modified data 6. **Use Appropriate Field Types**: Choose the right field type for better data validation and UI experience 7. **Cross-Team Collaboration**: Create shared bases for data that multiple AI teams need to access