Oracle Database MCP Server

# MCP Generator Examples Ready-to-use prompts for common MCP server scenarios. Just copy, customize, and paste to your AI assistant. ## 📚 Important: Always Fetch Latest Documentation **Before using these examples, instruct your AI assistant to fetch:** - MCP Specification: https://spec.modelcontextprotocol.io/ - MCP TypeScript SDK: https://github.com/modelcontextprotocol/typescript-sdk - MCP Servers Examples: https://github.com/modelcontextprotocol/servers These examples provide quick templates but the official documentation is the source of truth. --- ## 🗄️ Example 1: PostgreSQL MCP Server **Use Case:** Query PostgreSQL databases with AI assistance **Copy this prompt:** ``` Create a standalone MCP server for PostgreSQL databases on macOS. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-postgresql - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-postgresql-server - Purpose: Enable AI assistants to query PostgreSQL databases - Data Source: PostgreSQL database via pg library TOOLS TO CREATE: 1. query_database - Execute SELECT queries with timeout and row limits 2. get_database_schema - List tables or get column info for specific table 3. explain_query - Get PostgreSQL EXPLAIN plan for optimization ARCHITECTURE: Follow MCP best practices with: - src/database/postgresConnection.ts - Connection pool management - src/database/queryExecutor.ts - Query execution with safety limits - src/tools/ - Three tools above - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - Environment config for: PG_HOST, PG_PORT, PG_DATABASE, PG_USER, PG_PASSWORD - Connection pooling (min 2, max 10) - Query timeout: 30 seconds - Max rows: 1000 - TypeScript ES2022 modules (.js imports) - MCP-compliant response format DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - pg@^8.11.0 - @types/pg@^8.10.0 - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 Generate complete project structure and provide initialization commands for macOS terminal. ``` --- ## 🐙 Example 2: GitHub MCP Server **Use Case:** Interact with GitHub repositories and issues **Copy this prompt:** ``` Create a standalone MCP server for GitHub API integration on macOS/VS Code. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-github - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-github-server - Purpose: Query GitHub repositories, issues, and pull requests via AI - Data Source: GitHub REST API v3 via Octokit TOOLS TO CREATE: 1. search_repositories - Search repos by query, language, stars 2. get_repository_info - Get details about a specific repository 3. list_issues - List issues for a repository with filters 4. get_pull_request - Get PR details including reviews and status ARCHITECTURE: Follow MCP best practices with: - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - src/github/client.ts - Octokit client singleton - src/github/types.ts - GitHub API types - src/tools/ - Four tools above - Environment config for: GITHUB_TOKEN, GITHUB_API_URL - Rate limiting: 60 requests/minute - Request timeout: 10 seconds - TypeScript ES2022 modules (.js imports) - MCP-compliant response format DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - @octokit/rest@^20.0.0 - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 Generate complete project structure and provide initialization commands for macOS terminal. ``` --- ## 📁 Example 3: File System MCP Server **Use Case:** Search and analyze local files **Copy this prompt:** ``` Create a standalone MCP server for file system operations on macOS. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-filesystem - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-filesystem-server - Purpose: Search, read, and analyze files on local system - Data Source: macOS file system via Node.js fs API TOOLS TO CREATE: 1. search_files - Search files by name pattern in directory (recursive option) 2. read_file - Read file contents with encoding options 3. get_file_stats - Get file metadata (size, dates, permissions) 4. search_content - Grep-like search within files ARCHITECTURE: Follow MCP best practices with: - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - src/filesystem/operations.ts - Core file operations - src/filesystem/types.ts - File system types - src/tools/ - Four tools above - Environment config for: ALLOWED_DIRECTORIES (security), MAX_FILE_SIZE - Security: Only access allowed directories - Max file size: 10MB - Search depth limit: 5 levels - TypeScript ES2022 modules (.js imports) - MCP-compliant response format DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 - glob@^10.3.0 Generate complete project structure with security considerations and provide initialization commands for macOS terminal. ``` --- ## 🔔 Example 4: Slack MCP Server **Use Case:** Send messages and query Slack workspaces **Copy this prompt:** ``` Create a standalone MCP server for Slack integration on macOS/VS Code. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-slack - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-slack-server - Purpose: Send messages and query Slack channels via AI - Data Source: Slack Web API via @slack/web-api TOOLS TO CREATE: 1. send_message - Post message to channel or user 2. list_channels - List all channels in workspace 3. search_messages - Search messages by query and filters 4. get_user_info - Get user profile information ARCHITECTURE: Follow MCP best practices with: - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - src/slack/client.ts - WebClient singleton - src/slack/types.ts - Slack API types - src/tools/ - Four tools above - Environment config for: SLACK_TOKEN, SLACK_WORKSPACE_ID - Rate limiting: 1 request per second - Message length limit: 4000 characters - TypeScript ES2022 modules (.js imports) - MCP-compliant response format DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - @slack/web-api@^6.11.0 - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 Generate complete project structure and provide initialization commands for macOS terminal. ``` --- ## 📊 Example 5: MongoDB MCP Server **Use Case:** Query MongoDB collections **Copy this prompt:** ``` Create a standalone MCP server for MongoDB databases on macOS. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-mongodb - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-mongodb-server - Purpose: Query MongoDB collections via AI assistance - Data Source: MongoDB via official mongodb driver TOOLS TO CREATE: 1. query_collection - Find documents with MongoDB query syntax 2. get_schema - Infer schema from collection samples 3. aggregate - Run aggregation pipelines 4. list_collections - List all collections in database ARCHITECTURE: Follow MCP best practices with: - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - src/database/mongoConnection.ts - MongoClient connection manager - src/database/queryExecutor.ts - Query execution with limits - src/tools/ - Four tools above - Environment config for: MONGO_URI, MONGO_DATABASE - Connection pooling - Query timeout: 30 seconds - Max documents: 1000 - TypeScript ES2022 modules (.js imports) - MCP-compliant response format DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - mongodb@^6.3.0 - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 Generate complete project structure and provide initialization commands for macOS terminal. ``` --- ## 🌐 Example 6: REST API MCP Server **Use Case:** Generic REST API wrapper **Copy this prompt:** ``` Create a standalone MCP server for generic REST API integration on macOS. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-rest-api - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-rest-api-server - Purpose: Make HTTP requests to any REST API via AI - Data Source: HTTP/HTTPS endpoints via axios TOOLS TO CREATE: 1. http_get - Make GET request with headers and query params 2. http_post - Make POST request with body and headers 3. http_put - Make PUT request for updates 4. http_delete - Make DELETE request ARCHITECTURE: Follow MCP best practices with: - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - src/api/client.ts - Axios instance with interceptors - src/api/types.ts - HTTP types - src/tools/ - Four tools above - Environment config for: API_BASE_URL, API_KEY, API_TIMEOUT - Request timeout: 30 seconds - Response size limit: 5MB - Support for JSON and form data - TypeScript ES2022 modules (.js imports) - MCP-compliant response format DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - axios@^1.6.0 - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 Generate complete project structure and provide initialization commands for macOS terminal. ``` --- ## 📈 Example 7: Analytics/Metrics MCP Server **Use Case:** Query application metrics and logs **Copy this prompt:** ``` Create a standalone MCP server for application analytics on macOS. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-analytics - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-analytics-server - Purpose: Query and analyze application metrics and logs - Data Source: Log files and metrics databases (InfluxDB or similar) TOOLS TO CREATE: 1. query_metrics - Query time-series metrics with aggregations 2. search_logs - Search application logs by level, timestamp, message 3. get_error_stats - Aggregate error statistics 4. analyze_performance - Get performance metrics for time range ARCHITECTURE: Follow MCP best practices with: - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - src/analytics/client.ts - Metrics database client - src/analytics/logParser.ts - Log parsing utilities - src/tools/ - Four tools above - Environment config for: METRICS_URL, LOGS_DIR, RETENTION_DAYS - Query timeout: 60 seconds - Max metrics points: 10000 - TypeScript ES2022 modules (.js imports) - MCP-compliant response format DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - @influxdata/influxdb-client@^1.33.0 (or similar) - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 Generate complete project structure and provide initialization commands for macOS terminal. ``` --- ## 🛠️ Customization Template Use this template for your own custom MCP server: ``` Create a standalone MCP server for [YOUR USE CASE] on macOS/VS Code. BEFORE STARTING: Fetch and review latest MCP documentation: - https://spec.modelcontextprotocol.io/ - https://github.com/modelcontextprotocol/typescript-sdk PROJECT SETUP: - Directory: ~/projects/mcp-[name] - Please instruct me to create this directory if it doesn't exist PROJECT DETAILS: - Name: mcp-[name]-server - Purpose: [What it does] - Data Source: [Database/API/Service] TOOLS TO CREATE: 1. [tool1_name] - [description] 2. [tool2_name] - [description] 3. [tool3_name] - [description] ARCHITECTURE: Follow MCP best practices with: - src/server.ts - Main MCP server with stdio transport - src/client.ts - Test client - src/config.ts - Zod-validated environment config - src/[datasource]/[connection-file].ts - [Description] - src/[datasource]/types.ts - Type definitions - src/tools/ - Tool implementations - Environment config for: [LIST_ENV_VARS] - TypeScript ES2022 modules (.js imports) - MCP-compliant response format - [Any special considerations: timeouts, limits, pooling] DEPENDENCIES: - @modelcontextprotocol/sdk@^1.20.2 - [your specific client library] - zod@^3.25.76 - dotenv@^16.3.1 - typescript@^5.3.3 Generate complete project structure and provide initialization commands for macOS terminal. ``` --- ## 💡 Tips for Best Results 1. **Be specific** about your data source and operations 2. **List exact tools** you want to create (2-4 is ideal to start) 3. **Include environment variables** needed for your service 4. **Mention any special requirements** (rate limits, security, etc.) 5. **Specify initialization directory** for macOS 6. **Request initialization commands** to be included ## 🔗 Related Files - **Detailed Guide:** See `MCP-PROJECT-GENERATOR-PROMPT.md` - **Quick Start:** See `QUICK-START-GENERATOR.md` - **Official MCP Docs:** https://modelcontextprotocol.io/ - **MCP Specification:** https://spec.modelcontextprotocol.io/ - **MCP TypeScript SDK:** https://github.com/modelcontextprotocol/typescript-sdk --- **Ready to build? Copy one of the examples above and start generating! 🚀**