NewCrm MCP Server Foundation
Read-only MCP foundation server for NewCrm CRM, exposing a single /mcp endpoint and currently supporting GET /api/leads via the existing backend API without direct MongoDB access.
README
NewCrm MCP Server Foundation
This project is a separate MCP server for the existing NewCrm CRM. It does not modify the CRM codebase and does not connect directly to MongoDB.
Current scope is intentionally limited to a clean foundation for future MCP integration.
Architecture
ChatGPT Work ↓ OAuth 2.1 ↓ NewCrm MCP Server ↓ Existing NewCrm Backend APIs ↓ MongoDB
The MCP server is designed to communicate with the existing CRM through the existing backend API layer wherever possible.
Current status
The project intentionally does not include:
- OAuth implementation
- MCP tools
- HubSpot integration
- write operations
- create, update, or delete operations
- MongoDB access
- bulk exports
- webhooks
- batch updates
Project structure
NewCrm-MCP/
├── src/
│ ├── audit/
│ │ └── logger.js
│ ├── auth/
│ │ └── tokenValidator.js
│ ├── config/
│ │ └── env.js
│ ├── permissions/
│ │ └── roleDefinitions.js
│ ├── services/
│ │ └── crmApiClient.js
│ ├── tools/
│ │ └── index.js
│ └── server.js
├── .env.example
├── .gitignore
├── package.json
├── README.md
└── node_modules/
Installation
- Open a terminal in the project folder.
- Install dependencies:
npm install
Environment variables
Copy the example file and set values in your local .env file:
copy .env.example .env
The .env file should contain values such as:
PORTHOSTNODE_ENVCRM_API_BASE_URL=https://www.jtcrm.in/apiMCP_SERVER_NAMEMCP_SERVER_VERSIONLOG_LEVELCRM_API_TIMEOUT_MS
Do not commit secrets or real tokens to source control. Do not store a permanent admin JWT or any production credential in the project.
CRM API integration
This project includes a reusable CRM API client for safe read-only access to the existing NewCrm backend APIs.
Current supported endpoint:
GET /api/leads
The service is intentionally restricted to explicitly configured endpoints only. It does not accept arbitrary URLs from MCP users.
Example usage:
import { getLeads } from "./src/services/crmApiClient.js";
const result = await getLeads({
authToken: process.env.CRM_API_TEST_AUTH_TOKEN,
});
This keeps the integration reusable for future CRM API calls while avoiding direct MongoDB access and write operations.
Safe testing guidance
For development or testing, pass a user-scoped JWT at runtime from the caller or a local ephemeral environment value. Do not put a shared admin token in .env.
Example local test flow:
$env:CRM_API_TEST_AUTH_TOKEN="<user_jwt_here>"
node --input-type=module -e "import { getLeads } from './src/services/crmApiClient.js'; const result = await getLeads({ authToken: process.env.CRM_API_TEST_AUTH_TOKEN }); console.log(result);"
The service will never log the token itself and only sends the x-auth-token header when a caller provides it.
Run the server
npm start
Optional development mode:
npm run dev
How /mcp works
This project uses the official MCP SDK with Streamable HTTP transport.
The server exposes a single MCP endpoint at /mcp.
POST /mcphandles MCP JSON-RPC initialization and follow-up requests.- A new session is created when an initialization request is received.
- Each session stores its associated Streamable HTTP transport.
DELETE /mcpcloses a session when it is no longer needed.GET /mcpcurrently returns405 Method Not Allowedbecause this foundation does not yet support streaming GET requests.
This is the correct foundation for future tool and backend integrations without adding CRM functionality prematurely.
Testing the server
After starting the server, confirm it is listening successfully:
curl -i http://127.0.0.1:3002/mcp
Expected behavior:
- The server should respond with HTTP
405for GET requests on/mcp. - The server should also accept POST requests for MCP initialization and JSON-RPC traffic once a compatible client is connected.
A successful startup is indicated by logs similar to:
{"timestamp":"...","level":"info","message":"NewCrm MCP foundation server started","host":"127.0.0.1","port":3002,"endpoint":"/mcp"}
Notes
This repository is intentionally isolated from the existing NewCrm project. It is a stand-alone MCP foundation server and does not include direct database access or CRM-specific business logic.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.