Skip to main content
Glama

Fusion MCP Server

by FallseF
WARP.md3.93 kB
# WARP.md This file provides guidance to WARP (warp.dev) when working with code in this repository. ## Project Overview Fusion MCP Server is a Model Context Protocol (MCP) server that provides enhanced AI interactions through fusion algorithms and data transformation capabilities. It implements the MCP specification to expose AI tools for data analysis and transformation. ## Core Architecture The server follows a single-file architecture pattern with clear separation of concerns: - **FusionMCPServer Class**: Main server implementation that handles MCP protocol communication - **Tool Registration**: Uses MCP SDK's request handler pattern for tool discovery and execution - **Transport Layer**: Communicates via stdio using StdioServerTransport - **Tool Handlers**: Individual methods for each tool (`fusion_analyze`, `fusion_transform`) ### Key Components - `src/index.js` - Main entry point containing the complete server implementation - Tool handlers are implemented as async methods within the FusionMCPServer class - JSON schema validation is handled through MCP SDK's built-in schemas - All communication follows MCP protocol standards for request/response format ## Development Commands ### Basic Operations ```bash # Install dependencies npm install # Start the server in production mode npm start # Start development server with auto-reload npm run dev # Run tests npm test # Run linting npm run lint # Format code npm run format ``` ### Development Workflow The server uses Node.js's built-in `--watch` flag for development, eliminating the need for external tools like nodemon. The `dev` script automatically restarts the server when source files change. ## MCP Tool Implementation The server exposes two main tools: ### fusion_analyze - **Purpose**: Analyze data using statistical, ML, or hybrid methods - **Input Schema**: `data` (required), `method` (optional: "statistical", "ml", "hybrid") - **Implementation**: Located in `handleAnalyze()` method ### fusion_transform - **Purpose**: Transform data between different formats using fusion techniques - **Input Schema**: `input` (required), `target_format` (required) - **Implementation**: Located in `handleTransform()` method ### Adding New Tools To add new MCP tools: 1. Add tool definition to the `ListToolsRequestSchema` handler array 2. Add case statement to `CallToolRequestSchema` handler switch 3. Implement handler method following the pattern: `async handleToolName(args)` 4. Return response in MCP format with `content` array containing `type: "text"` objects ## Technical Requirements - **Node.js**: >= 18.0.0 (specified in package.json engines) - **Module System**: ES modules (`"type": "module"` in package.json) - **MCP SDK**: Uses `@modelcontextprotocol/sdk` version ^0.4.0 - **Development Tools**: ESLint ^8.57.0, Prettier ^3.0.0 ## Testing Strategy The project uses Node.js's built-in test runner (`node --test`). Tests should be created following Node.js test conventions and placed appropriately within the project structure. ## Code Standards - **ES Modules**: All imports/exports use ES module syntax - **Async/Await**: Consistent use of async/await for asynchronous operations - **Error Handling**: Proper error throwing for unknown tools and invalid requests - **JSON Schema**: All tool inputs validated through MCP SDK schema validation - **Code Formatting**: Prettier configuration for consistent code style - **Linting**: ESLint for code quality and style enforcement ## MCP Protocol Compliance The server implements core MCP protocol features: - Tool listing via `ListToolsRequestSchema` - Tool execution via `CallToolRequestSchema` - Proper request/response format with content arrays - Error handling for unknown tools - Stdio transport for communication When extending functionality, maintain compliance with MCP specification standards for request/response formats and tool schema definitions.

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/FallseF/Fusion_MCP'

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