salesforce-mcp
# MCP Salesforce Connector
A Model Context Protocol (MCP) server implementation for Salesforce integration, allowing LLMs to interact with Salesforce data through SOQL queries and SOSL searches.
## Features
- Execute SOQL (Salesforce Object Query Language) queries
- Perform SOSL (Salesforce Object Search Language) searches
- Retrieve metadata for Salesforce objects, including field names, labels, and types
- Retrieve, create, update, and delete records
- Execute Tooling API requests
- Execute Apex REST requests
- Make direct REST API calls to Salesforce
## Configuration
### Model Context Protocol
To use this server with the Model Context Protocol, you need to configure it in your `claude_desktop_config.json` file. Add the following entry to the `mcpServers` section:
{
"mcpServers": {
"salesforce": {
"command": "uvx",
"args": [
"--from",
"mcp-salesforce-connector",
"salesforce"
],
"env": {
"SALESFORCE_USERNAME": "YOUR_SALESFORCE_USERNAME",
"SALESFORCE_PASSWORD": "YOUR_SALESFORCE_PASSWORD",
"SALESFORCE_SECURITY_TOKEN": "YOUR_SALESFORCE_SECURITY_TOKEN"
}
}
}
}
Replace `YOUR_SALESFORCE_USERNAME`, `YOUR_SALESFORCE_PASSWORD`, and `YOUR_SALESFORCE_SECURITY_TOKEN` with your Salesforce credentials.
TDQS
Scored across 10 tools
Most tools have clear distinct purposes, but there is some functional overlap that could cause confusion. For example, apex_execute, restful, and tooling_execute all involve making API calls with different endpoints, which might be ambiguous without careful reading of descriptions. The CRUD operations (create_record, get_record, update_record, delete_record) are well-differentiated from query/search tools (run_soql_query, run_sosl_search), but the multiple execution methods create some redundancy.
Tool names follow a mostly consistent verb_noun pattern with clear actions like create_record, delete_record, get_record, and update_record. However, there are minor deviations such as apex_execute and tooling_execute using 'execute' instead of a more specific verb, and restful being a noun rather than a verb_noun combination. Overall, the naming is readable and predictable with only slight inconsistencies.
With 10 tools, the count is well-scoped for a Salesforce integration server. It covers essential operations like CRUD, queries, searches, and API executions without being overwhelming. Each tool appears to serve a distinct purpose in the Salesforce domain, making the set manageable and comprehensive for typical agent tasks.
The tool set provides strong coverage of core Salesforce operations, including record management (create, read, update, delete), data querying (SOQL, SOSL), and API interactions (REST, Apex, Tooling). A minor gap is the lack of tools for metadata operations beyond get_object_fields, such as listing objects or managing schema changes, but agents can likely work around this using existing tools like restful or tooling_execute.