salesforce_mcp_server

sitrakaherivalisoagithub/salesforce_mcp_server

3.2

If you are the rightful owner of salesforce_mcp_server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The Salesforce MCP Server is a bridge between a Large Language Model (LLM) and the Salesforce API, facilitating interaction with Salesforce data.

Tools
9
Resources
0
Prompts
0

Salesforce MCP Server

This project provides a Model-Context-Protocol (MCP) server that acts as a bridge between a Large Language Model (LLM) and the Salesforce API. It exposes a set of tools that allow the LLM to interact with Salesforce data, covering both Sales Cloud and Service Cloud functionalities.

Features

  • Salesforce Integration: Connects securely to a Salesforce instance using user credentials.
  • Tool Exposure: Exposes a rich set of tools for interacting with Salesforce objects like Contacts, Accounts, Opportunities, and Cases.
  • Sales Cloud Tools: Search contacts, get account details, list open opportunities, and log activities.
  • Service Cloud Tools: Get case details, list open cases, create new cases, and add comments.
  • Global Search: Perform a global search across multiple Salesforce objects.
  • MCP Compliant: Built on the mcp.server framework for seamless integration with MCP-compatible clients.

Prerequisites

To run this server, you need valid Salesforce credentials. The server can be configured via environment variables or command-line arguments.

Environment Variables:

  • SF_USERNAME: Your Salesforce username.
  • SF_PASSWORD: Your Salesforce password.
  • SF_TOKEN: Your Salesforce security token.
  • SF_DOMAIN: The Salesforce domain to use. Defaults to login for production. Use test for sandboxes.

Installation

  1. Clone the repository:

    git clone <repository-url>
cd salesforce_mcp_server

  2. Install dependencies: This project uses uv for package management.

    uv pip install -r requirements.txt 
# Or if you have the pyproject.toml and uv.lock
uv sync

    (Note: A requirements.txt may need to be generated from pyproject.toml if not present)

Running the Server

You can start the server using the run.py script. You must provide your Salesforce credentials either as environment variables (see Prerequisites) or as command-line arguments.

python run.py --sf-username "your_user@example.com" --sf-password "your_password" --sf-token "your_token"

Command-line options:

  • --port: Port to run the server on (default: 3000).
  • --host: Host to bind to (default: 127.0.0.1).
  • --sf-username: Salesforce username.
  • --sf-password: Salesforce password.
  • --sf-token: Salesforce security token.
  • --sf-domain: Salesforce domain (default: login).
  • --log-level: Logging level (e.g., INFO, DEBUG).

Available Tools

The server exposes the following tools to the LLM:

Sales Cloud Tools

  • search_contact: Searches for a contact in Salesforce by name and optionally by email.
  • get_account_details: Retrieves the full details of a customer account by its exact name or ID.
  • list_open_opportunities: Lists all sales opportunities (deals) that are not yet closed. Can be filtered by account or owner.
  • log_activity: Records an activity (like a call or email) and associates it with a record (Contact, Account, Opportunity).

Service Cloud Tools

  • get_case_details: Retrieves the details of a support ticket (Case) using its ticket number.
  • list_open_cases: Lists all open support tickets, with optional filtering by contact email or account name.
  • create_case: Creates a new support ticket for a customer.
  • add_comment_to_case: Adds a comment (internal note or public reply) to an existing support ticket.

Global Search Tool

  • search_salesforce: Performs a global search across Salesforce (Accounts, Contacts, Opportunities, Cases) using a search term.