mcp
A Model Context Protocol Server facade to simplify the implementation of agents
claude mcp add --transport stdio agentico-dev-mcp-server node build/index.js
How to use
This MCP server implements the Model Context Protocol (MCP) to provide a simplified API for interacting with tools via a server that you can extend with your own Tool implementations. It ships with a facade pattern that lets you register custom tools and expose them through the MCP API, making it easier for clients to discover and execute those tools. The README example shows how to implement a basic EchoTool as a demonstration: you create a Tool subclass, define its schema, and register it with the MCPServer. After building and starting the server, clients can invoke registered tools by name and pass the required input according to each tool's schema. The server architecture is designed to be extended by adding more tools in separate files and registering them at startup.
How to install
Prerequisites:
- Node.js and npm/yarn installed
- TypeScript and build setup if you plan to develop tools
Step-by-step installation:
- Clone the repository or create your project that will host the MCP server.
- Install dependencies (adjust to your package manager):
- npm install or yarn install
- If you are starting from the example in the README, add the MCP server package to your project:
- yarn add @agentico/mcp-server
- Create your server entry (for example build/index.js after compiling TypeScript) and implement/register your tools as shown in the README example.
- Build the project (if using TypeScript):
- yarn build
- Run the server:
- yarn start
- or node build/index.js
Notes:
- Ensure your TypeScript transpilation outputs to build/index.js (or adjust the mcp_config path accordingly).
- You can test with MCP clients that support MCP once the server is running.
Additional notes
Tips and common considerations:
- If you add more tools, keep their schemas clear and well-documented so clients can construct correct inputs.
- The server is designed to be extended with your own tools under the tools/ directory or similar structure; follow the EchoTool example to implement custom behavior.
- When deploying, verify the runtime environment (Node version, package externals) matches your build outputs.
- If you encounter port or binding issues, consider configuring environment variables for the server (e.g., PORT) in your deployment setup, even though the README example does not explicitly define them.
- For debugging, run in development mode with console logs and catch blocks as demonstrated in the example main function.
Related MCP Servers
ext-apps
Official repo for spec & SDK of MCP Apps protocol - standard for UIs embedded AI chatbots, served by MCP servers
dexto
A coding agent and general agent harness for building and orchestrating agentic applications.
mcp-graphql
Model Context Protocol server for GraphQL
toolsdk -registry
MCPSDK.dev(ToolSDK.ai)'s Awesome MCP Servers and Packages Registry and Database with Structured JSON configurations. Supports OAuth2.1, DCR...
mcp -typescript
DataForSEO API modelcontextprotocol server
mcp-gemini
This project provides a dedicated MCP (Model Context Protocol) server that wraps the @google/genai SDK. It exposes Google's Gemini model capabilities as standard MCP tools, allowing other LLMs (like Cline) or MCP-compatible systems to leverage Gemini's features as a backend workhorse.