
Orion Salesforce MCP Server
Enables natural language querying and data manipulation of Salesforce orgs, with intelligent SOQL/SOSL generation, cross-object search, and enterprise-ready features optimized for wholesale distribution systems.
README
Orion Salesforce MCP Server
A powerful and intelligent Salesforce data interaction server designed specifically for wholesale distribution systems. This MCP (Model Control Plane) server provides natural language querying capabilities with advanced SOQL/SOSL generation, intelligent field resolution, and comprehensive data manipulation features.
🌟 Key Features
Intelligent Data Querying
- Natural Language to SOQL: Convert plain English questions into optimized Salesforce queries
- Smart Object & Field Resolution: Automatically map business terms to Salesforce API names using configurable synonyms
- Cross-Object Search (SOSL): Search across multiple objects simultaneously with intelligent result aggregation
- Relationship Traversal: Automatically handle lookups and master-detail relationships
- Time-Sensitive Filtering: Intelligent date range detection and application
Data Manipulation (DML)
- Controlled DML Operations: Configurable Insert, Update, Upsert, Delete, Undelete, and Merge capabilities
- Feature Flags: Granular control over DML operations with org-specific configuration
- Safety Guardrails: Built-in limits, confirmation requirements, and bulk operation controls
- Permission Respect: Honor Salesforce field-level security and object permissions
Hybrid AI Responses
- Data-Driven Answers: Combine actual Salesforce data with LLM knowledge for contextual responses
- Intelligent Fallbacks: Graceful handling when data is unavailable or insufficient
- Business Context: Understand wholesale distribution terminology and processes
Enterprise-Ready Architecture
- Multi-Org Support: Configure different Salesforce orgs with unique settings and permissions
- Robust Error Handling: Comprehensive Salesforce API error handling with clear user feedback
- Security First: Token management, PII redaction, and configurable access controls
- Performance Optimized: Intelligent caching, query optimization, and efficient API usage
🚀 Quick Start
Prerequisites
- Node.js 18+
- Salesforce org with API access
- Valid Salesforce access tokens
Installation
-
Clone and Install
git clone https://github.com/jasonjuela/orion-salesforce-mcp-server.git cd orion-salesforce-mcp-server npm install
-
Configure Salesforce Tokens
cp data/secrets/tokens.example.json data/secrets/tokens.json
Edit
data/secrets/tokens.json
with your Salesforce credentials:{ "default": { "access_token": "your_access_token_here", "instance_url": "https://your-instance.salesforce.com", "token_type": "Bearer" } }
-
Customize Organization Profile
Edit
data/configs/default.json
to match your Salesforce org structure:{ "orgId": "your-org-name", "namespaces": ["your_namespace__"], "objectSynonyms": { "owsc__Item__c": ["wine", "product", "inventory"], "owsc__Order__c": ["order", "purchase order", "po"] } }
-
Start the Server
npm start
Server runs on
http://localhost:3018
📖 Usage Examples
Basic Querying
# Natural language queries
curl -X POST http://localhost:3018/generate/stream \
-H "Content-Type: application/json" \
-d '{
"user_question": "What is the alcohol percentage of Cockburn'\''s wines?",
"orgId": "default",
"sessionId": "demo"
}'
Cross-Object Search
# Search across multiple objects
curl -X POST http://localhost:3018/generate/stream \
-H "Content-Type: application/json" \
-d '{
"user_question": "search for Smith across all objects",
"orgId": "default",
"sessionId": "demo"
}'
Test Interface
Open http://localhost:3018
in your browser for a simple web interface to test queries.
⚙️ Configuration
Organization Profiles (data/configs/default.json
)
{
"orgId": "wholesale-system",
"description": "Wholesale distribution system configuration",
"namespaces": ["owsc__"],
"guardrails": {
"piiRedaction": true,
"maxRows": 1000,
"defaultDateRange": "LAST_N_MONTHS:6",
"allowedObjects": [],
"blockedObjects": ["UserRecordAccess", "ObjectPermissions"]
},
"features": {
"dmlOperations": {
"enabled": false,
"allowedOperations": {
"insert": false,
"update": false,
"upsert": false,
"delete": false,
"undelete": false,
"merge": false
},
"safeguards": {
"maxRecordsPerOperation": 50,
"requireConfirmation": true,
"allowBulkOperations": false
}
}
},
"objectSynonyms": {
"owsc__Item__c": ["wine", "product", "inventory", "item"],
"owsc__Order__c": ["order", "purchase order", "po"],
"Account": ["account", "customer", "client"]
}
}
DML Operations
To enable data manipulation:
- Set
features.dmlOperations.enabled: true
- Enable specific operations (
insert
,update
, etc.) - Configure safety guardrails
- Specify allowed/blocked objects in
dmlRestrictions
🏗️ Architecture
Core Components
- Planner Service (
src/services/planner.js
): Intent detection and query generation - Salesforce Service (
src/services/salesforce.js
): API client with retry logic - Intelligent Resolver (
src/services/intelligentResolver.js
): Object and field mapping - Stream Generator (
src/routes/generateStream.js
): Main orchestration and LLM integration
Data Flow
- Intent Detection: Analyze user question for search type and entities
- Object Resolution: Map business terms to Salesforce objects using synonyms
- Field Discovery: Use Describe API to find relevant fields
- Query Generation: Build optimized SOQL/SOSL queries
- Data Retrieval: Execute queries with error handling and retries
- Response Generation: Combine data with LLM for intelligent answers
🔧 API Endpoints
Primary Endpoints
POST /generate/stream
- Main query interface with streaming responsesGET /auth/login
- Salesforce OAuth initiation (placeholder)GET /auth/callback
- OAuth callback handler (placeholder)GET /search/capabilities
- Check available search capabilitiesPOST /search/cross-object
- Direct SOSL search endpoint
Response Format
Streaming responses include:
- Query plan (SOQL/SOSL)
- Data retrieval status
- Record count
- Formatted results
- LLM-generated insights
🛡️ Security Features
- Token Management: Secure credential storage and rotation
- PII Redaction: Configurable personally identifiable information filtering
- Object Whitelisting: Control access to specific Salesforce objects
- Field-Level Security: Respect Salesforce permissions
- Query Limits: Prevent expensive or dangerous operations
- Feature Flags: Granular control over dangerous operations like DML
🎯 Wholesale Distribution Focus
Optimized for wholesale/distribution use cases:
- Inventory Management: Track items, lots, containers, and locations
- Order Processing: Handle purchase orders, sales orders, and fulfillment
- Supply Chain: Manage vendors, warehouses, and shipping
- Financial Operations: Support invoicing, payments, and accounting
- Business Intelligence: Aggregate data for reporting and analytics
🚦 Production Considerations
Performance
- Implement response caching for frequent queries
- Use connection pooling for Salesforce API calls
- Add request rate limiting
- Monitor API usage limits
Security
- Implement proper OAuth flows
- Add request authentication/authorization
- Enable HTTPS in production
- Regular token rotation
- Audit logging
Monitoring
- Add comprehensive logging
- Implement health checks
- Monitor Salesforce API limits
- Track query performance metrics
🤝 Contributing
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature
) - Commit changes (
git commit -m 'Add amazing feature'
) - Push to branch (
git push origin feature/amazing-feature
) - Open a Pull Request
📄 License
This project is licensed under the MIT License - see the LICENSE file for details.
🆘 Support
For questions, issues, or feature requests:
- Open an issue on GitHub
- Check existing documentation
- Review configuration examples
Built for intelligent Salesforce data interaction with wholesale distribution in mind.
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.