MCP ServerHTTPOfficialv1.0.0

Smithery Kodey AI Salesforce MCP Server

Query and manage Salesforce orgs through SOQL and metadata tools via LLM-driven chat. Built for Salesforce developers who want AI-assisted data exploration, record inspection, and deployment workflows without exposing secrets.

ai.smithery/kodey-ai-salesforce-mcp

Hosted URL

https://server.smithery.ai/@kodey-ai/salesforce-mcp/mcp

Transport

HTTP

Auth

No auth required

Smithery Kodey AI Salesforce repository at a glance

Live signal from GitHub, refreshed weekly.

Stars

0

Last commit

Oct 24, 2025

License

Apache-2.0

Language

JavaScript

What the Smithery Kodey AI Salesforce MCP server does

How models use it and what it is built for.

Query and manage Salesforce orgs through SOQL and metadata tools via LLM-driven chat. Built for Salesforce developers who want AI-assisted data exploration, record inspection, and deployment workflows without exposing secrets.

Connect to Smithery Kodey AI Salesforce

Hosted endpoint — paste into any MCP client.

https://server.smithery.ai/@kodey-ai/salesforce-mcp/mcp

Resources

Where to find authoritative docs and source for Smithery Kodey AI Salesforce.

Example prompts for Smithery Kodey AI Salesforce

Paste any of these into Agent Studio after connecting Smithery Kodey AI Salesforce.

  • Show me all accounts in my org with alias production-org
  • List the fields on the Contact object and their data types
  • Query all opportunities with amount greater than 50000 in the last quarter
  • Deploy my project metadata to the dev-org and report any errors

Documentation from project README

View on GitHub

Excerpted from the project's README — boilerplate sections (license, changelog, contributing) omitted for clarity.

mcp

MCP Server for Interacting with Salesforce Orgs

NPM License smithery badge

Feedback

Report bugs and issues here.
For feature requests and other related topics, start a Discussion here.

Overview of the Salesforce DX MCP Server (Developer Preview)

The Salesforce DX MCP Server is a specialized Model Context Protocol (MCP) implementation designed to facilitate seamless interaction between large language models (LLMs) and Salesforce orgs. This MCP server provides a robust set of tools and capabilities that enable LLMs to read, manage, and operate Salesforce resources securely.

Key Features:

  • Direct interaction with Salesforce orgs through LLM-driven tools.
  • Secure access using TypeScript libraries (not shelling out to the sf Salesforce CLI).
  • Improved security by avoiding the exposure of secrets in plain text.
  • Granular access control with org allowlisting.
  • Modular tool architecture for easy extensibility.

NOTE: The Salesforce DX MCP Server is available as a developer preview. The feature isn’t generally available unless or until Salesforce announces its general availability in documentation or in press releases or public statements. All commands, parameters, and other features are subject to change or deprecation at any time, with or without notice. Don't implement functionality developed with these commands or tools. As we continue to enhance and refine the implementation, the available functionality and tools may evolve. We welcome feedback and contributions to help shape the future of this project.

Security Features

The Salesforce DX MCP Server was designed with security as a top priority.

  • Uses TypeScript libraries directly

    • Greatly decreases the size of the MCP Server.
    • Significantly reduces the risk of remote code execution (RCE).
  • No secrets needed in configuration

    • Eliminates the risk of plain text secret exposure.
    • Accesses pre-existing (encrypted) auth files on the user's machine.
    • Implements allowlisting for auth info key/values to prevent sensitive data exposure.
  • No secrets exposed via MCP tools

    • Prevents other tools from accessing unencrypted tokens.
    • Tools pass usernames around instead of tokens.
  • Granular access control

    • MCP Server can access auth info for only orgs that have been explicitly allowlisted.
    • Users specify allowed orgs when starting the server.

Agentforce for Developers Includes the Salesforce DX MCP Server

Agentforce for Developers is an AI-powered Salesforce developer tool that's available as an easy-to-install Visual Studio Code (VS Code) extension. It includes Dev Agent (developer preview), an intelligent coding partner that provides information and, most importantly, can take action. Through agentic chat powered by MCP, Dev Agent can execute commands and perform complex workflows automatically, right from within VS Code.

The Salesforce DX MCP server is pre-configured in Agentforce for Developers, so you can start using the DX MCP tools immediately after you install the extension in VS Code.

See Set Up Agentforce for Developers and Chat with Dev Agent (Dev Preview) for more information.

Get Started Using VS Code as the Client

Want to jump in and see what all the fuss is about? Read on!

This example uses Visual Studio Code (VS Code) as the MCP client. After you configure it with the Salesforce DX MCP Server, you then use GitHub Copilot and natural language to easily execute typical Salesforce DX development tasks, such as listing your authorized orgs, viewing org records, and deploying or retrieving metadata.

But you're not limited to using only VS Code and Copilot as the MCP client! As already mentioned, you can use Agentforce for Developers, which is pre-configured with the DX MCP server. Or you can configure many other clients to use the Salesforce DX MCP Server, such as Cursor, Cline, Claude Desktop, Zed, Windsurf, and more.

Before You Begin

For the best getting-started experience, make sure that you have a Salesforce DX environment set up on your computer. In particular:

Let's Do It

  1. Create a .vscode/mcp.json file at the root of your DX project and add this JSON:

    {
      "servers": {
        "Salesforce DX": {
          "type": "stdio",
          "command": "npx",
          "args": ["-y", "@salesforce/mcp", "--orgs", "DEFAULT_TARGET_ORG", "--toolsets", "all"]
        }
      }
    }
    

    You can also configure the MCP server globally by editing the VS Code settings.json file and adding a similar JSON snippet but contained in an mcp:servers section.

    The --orgs flag is required and specifies the authorized orgs you're allowing the MCP server to access. The --toolsets and --tools flags are used to specify which tools the server will start with. See Configure the DX MCP Server for the available values for these flags.

  2. Open VS Code, go to View -> Command Palette and enter MCP: List Servers.

    TIP: You can also get to the command palette by pressing press Ctrl+Shift+P (Windows or Linux) or Command-Shift-P (macOS).

  3. Click Salesforce DX, then Start Server.

    Check the Output tab for the server status.

  4. Run Chat: Open Chat (Agent) from the command palette to start a new GitHub Copilot chat session.

    Be sure your Copilot chat window is in Agent mode; if you're in Ask or Edit mode, use the little drop-down to switch.

  5. In the GitHub Copilot chat window, use natural language to explain what you want to do. The MCP server determines which configured tool to use, and then shows it to you along with other information. Click Continue to run the tool and see the results of your request.

    Try out these examples:

    • List all my orgs.
    • Which are my active scratch orgs?
    • Show me all the accounts in the org with alias my-org.
    • Deploy everything in my project to the org with alias my-org.
    • Do you see any performance or security issues with the Apex code in the MyApexClass.cls file?
    • I see that my Apex code violates the pmd:ApexCRUDViolation rule; can you give me more information about this rule?
  6. To stop, restart, or view the MCP server configuration, run the MCP: List Servers command, click Salesforce DX, then click the appropriate option.

Installing via Smithery

To install salesforce-mcp-kodey automatically via Smithery:

npx -y @smithery/cli install @kodey-ai/salesforce-mcp-kodey
Continue reading on GitHub

Smithery Kodey AI Salesforce MCP server — FAQ

Common questions about connecting and running Smithery Kodey AI Salesforce.

  • What authentication does this MCP server use?

    It uses pre-existing encrypted auth files from your local Salesforce CLI setup (via `sf org login web`). No secrets are stored in plain text or exposed through the MCP tools—the server passes usernames instead of tokens for security.

  • Do I need to install Salesforce CLI to use this?

    Yes. You must install Salesforce CLI and authorize at least one org locally before the MCP server can access it. The server reads your encrypted auth files and respects org allowlisting you configure.

  • Can I use this with clients other than VS Code?

    Yes. While the README highlights VS Code with GitHub Copilot, the server also works with Agentforce for Developers, Cursor, Cline, Claude Desktop, Zed, Windsurf, and other MCP-compatible clients.

  • What is the difference between --toolsets and --tools flags?

    The `--toolsets` flag enables groups of related tools (e.g., 'all' enables everything). The `--tools` flag lets you enable individual tools by name and can be combined with `--toolsets` for fine-grained control.

  • Is this production-ready or still experimental?

    This is a developer preview. Salesforce states that all commands, parameters, and features are subject to change or deprecation at any time without notice, so avoid implementing production systems on it until general availability is announced.

Run Smithery Kodey AI Salesforce across 60+ AI models, side-by-side

Connect Smithery Kodey AI Salesforce to Claude, GPT, Gemini, DeepSeek and 60+ AI models in MCP Agent Studio. Compare answers side-by-side, save reusable agent presets, share runs — all in your browser, no install required.

Open Agent Studio

Related servers

More on MCP Playground