Salesforce CLI MCP Server
Enables AI assistants to interact with Salesforce orgs through natural language commands, supporting org management, SOQL queries, metadata operations, testing, and more.
README
Salesforce CLI MCP Server
A Model Context Protocol (MCP) server that integrates with the Salesforce CLI, allowing AI assistants to interact with Salesforce orgs through natural language commands. Now with Web UI!
🆕 New Features
- 🌐 Web UI: Interactive web interface for easy Salesforce CLI operations
- 🔗 HTTP Server: RESTful API endpoints alongside MCP protocol
- 📊 Real-time Status: Live server health monitoring
- 🎯 Quick Actions: One-click shortcuts for common operations
- 📋 Copy Results: Easy result copying and formatting
Features
- Org Management: List and manage authorized Salesforce orgs
- Data Operations: Execute SOQL queries and import data
- Metadata Operations: Deploy and retrieve metadata
- Testing: Run Apex tests
- Package Management: Create and manage Salesforce packages
- Custom Commands: Execute any Salesforce CLI command
- Web Interface: User-friendly web UI for all operations
- HTTP API: RESTful endpoints for programmatic access
Prerequisites
Before using this MCP server, ensure you have:
- Salesforce CLI installed: Install Salesforce CLI
- Node.js (version 18 or higher)
- At least one authorized Salesforce org
Quick Start
1. Clone the repository
git clone https://github.com/timaw513-emergenit/salesforce-cli-mcp-server.git
cd salesforce-cli-mcp-server
2. Install dependencies
npm install
3. Build the project
npm run build
4. Choose your mode
Option A: Web UI Mode (Recommended)
Start the server with web interface:
npm run web
Then open your browser to: http://localhost:3000
Option B: STDIO Mode (for MCP clients)
Configure your MCP client as shown below.
🌐 Web UI Usage
The web interface provides an intuitive way to interact with your Salesforce orgs:
Features:
- 🚀 Interactive Forms: Fill out forms for each Salesforce CLI command
- 📊 Quick Actions: One-click shortcuts for common tasks
- 📋 Live Results: Real-time command execution with formatted output
- 📱 Responsive Design: Works on desktop, tablet, and mobile
- 🔄 Status Monitoring: Real-time server health indicators
Available Operations:
- List Orgs: View all authorized Salesforce orgs
- Execute SOQL: Run queries with syntax highlighting
- Custom Commands: Execute any Salesforce CLI command
- Deploy/Retrieve: Metadata operations with progress tracking
- Run Tests: Execute Apex tests with detailed results
Quick Actions:
- List Orgs: Instantly view all authorized orgs
- Query Accounts: Pre-filled SOQL query for Account records
- Org Info: Display current org details
API Endpoints
When running in HTTP mode, the server provides these endpoints:
Core Endpoints:
GET /- Web UI interfaceGET /health- Health checkGET /api/status- Detailed server statusGET /api/tools- Available tools listPOST /api/execute- Execute Salesforce CLI commandsPOST /mcp- MCP protocol endpoint
Example API Usage:
# Execute a SOQL query
curl -X POST http://localhost:3000/api/execute \
-H "Content-Type: application/json" \
-d '{"command": "data query --json --query \"SELECT Id, Name FROM Account LIMIT 5\""}'
# List all orgs
curl -X POST http://localhost:3000/api/execute \
-H "Content-Type: application/json" \
-d '{"command": "org list --json"}'
# Check server health
curl http://localhost:3000/health
MCP Client Configuration
For Claude Desktop
Add this configuration to your Claude Desktop MCP settings file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%/Claude/claude_desktop_config.json
STDIO Mode:
{
"mcpServers": {
"salesforce-cli": {
"command": "node",
"args": ["/absolute/path/to/salesforce-cli-mcp-server/build/index.js"]
}
}
}
HTTP Mode:
{
"mcpServers": {
"salesforce-cli-http": {
"command": "curl",
"args": ["-X", "POST", "http://localhost:3000/mcp"]
}
}
}
For VS Code with MCP Extension
Create a .vscode/mcp.json file in your project root:
{
"servers": {
"Salesforce CLI": {
"type": "stdio",
"command": "node",
"args": ["/absolute/path/to/salesforce-cli-mcp-server/build/index.js"]
}
}
}
Available Commands
NPM Scripts
# Development
npm run build # Build TypeScript
npm run dev # Watch mode for development
npm run clean # Clean build directory
# Running
npm run start # Start in STDIO mode
npm run http # Start HTTP server (with Web UI)
npm run web # Alias for http mode
npm run ui # Alias for http mode
# Testing
npm run inspector # Run MCP Inspector
npm run test # Build and run inspector
Available Tools
1. sf_org_list
List all authorized Salesforce orgs.
2. sf_data_query
Execute a SOQL query against a Salesforce org.
3. sf_project_deploy
Deploy metadata to a Salesforce org.
4. sf_project_retrieve
Retrieve metadata from a Salesforce org.
5. sf_apex_test_run
Run Apex tests in a Salesforce org.
6. sf_data_import
Import data from a CSV file into a Salesforce org.
7. sf_package_create
Create a new Salesforce package.
8. sf_custom_command
Execute any Salesforce CLI command.
Usage Examples
Via Web UI
- Open http://localhost:3000
- Select a tool from the sidebar
- Fill out the form parameters
- Click "Execute" and view results
Via Natural Language (MCP clients)
Once configured, you can use natural language with your AI assistant:
Org Management
- "List all my Salesforce orgs"
- "Show me the connection status of all orgs"
Data Queries
- "Query all accounts in my org"
- "Get all contacts where the email contains 'example.com'"
- "Run this SOQL query: SELECT Id, Name FROM Account LIMIT 10"
Metadata Operations
- "Deploy all metadata to my sandbox"
- "Retrieve the Account object metadata"
- "Deploy only the AccountController class"
Testing
- "Run all Apex tests in my org"
- "Run tests for the AccountTest class"
Data Import
- "Import the accounts.csv file as Account records"
Custom Commands
- "Run the command: org display --target-org myorg"
Development
Testing
Use the MCP Inspector to test your server:
npm run inspector
Development Mode
Run in development mode with auto-rebuild:
npm run dev
Web UI Development
For web UI development, start the HTTP server and make changes to files in the public/ directory:
npm run web
The server will serve static files from the public/ directory automatically.
Security Considerations
This MCP server executes Salesforce CLI commands directly on your system. Ensure that:
- You trust the AI assistant you're using
- You have proper backup procedures in place
- You understand the implications of the commands being executed
- You review any destructive operations before confirming
- The HTTP server (if used) is only accessible from trusted networks
Troubleshooting
Common Issues
-
"Command not found: sf"
- Ensure Salesforce CLI is installed and in your PATH
-
"No orgs found"
- Make sure you have authorized at least one Salesforce org using
sf org login web
- Make sure you have authorized at least one Salesforce org using
-
"Permission denied"
- Ensure the build files have execute permissions:
chmod +x build/*.js
- Ensure the build files have execute permissions:
-
"Module not found"
- Run
npm installto install dependencies - Run
npm run buildto compile TypeScript
- Run
-
"Port already in use"
- Change the port:
MCP_PORT=3001 npm run web
- Change the port:
-
Web UI not loading
- Ensure you've run
npm run buildfirst - Check that
public/index.htmlexists - Verify server is running:
curl http://localhost:3000/health
- Ensure you've run
Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Add tests if applicable
- Submit a pull request
License
MIT License - see LICENSE file for details.
🌟 Now with Web UI! Access your Salesforce CLI through a beautiful, intuitive web interface at http://localhost:3000
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。