apex_execute
Executes an Apex REST request
LABSCO SUMMARY
The query tools are written for a token budget rather than for a console.
run_soql_query returns CSV by default and its guidance pushes named fields and a LIMIT, get_record returns every field and says as much, and get_object_fields exists so the schema is fetched once instead of guessed at across retries. Below the typed tools sit tooling_execute, apex_execute and restful, which will reach anything the authenticated profile can — useful, and the reason this belongs on an org where that profile is scoped deliberately. The bulk tools take whole record lists, so a wrong filter is wrong at scale.
What it is
A server over a Salesforce org's REST, Tooling and Apex endpoints — queries, record CRUD, bulk operations and a raw REST escape hatch — in 14 tools.
What you get
- SOQL queries and SOSL searches run directly, returning CSV by default and JSON on request when nested structure matters
- Field discovery before query writing: a per-object list of name, label, type and updateable, and an enumeration of every standard and custom SObject in the org
- Single-record work by id — fetch, create, update, delete
- Bulk create, update and delete across a list of records of one SObject type
- Tooling API and Apex REST requests for the work the data endpoints do not cover
- A direct REST call taking method, path, params and body, which reaches anything else in the org's API surface
Requirements
A Salesforce org this machine can authenticate to, and uv with a Python available since the package runs under it. The profile behind that authentication is the boundary — the Apex and raw REST tools reach whatever it can. MIT.
Setup effort
One command plus credentials — uvx --from mcp-salesforce-connector salesforce, then supply credentials
USE CASES
01
Answering a pipeline question without exporting a report
Read the object's real field list first — names, types and whether each is updateable — then write a SOQL query that asks for only those fields and returns CSV.
02
Finding a record when you only remember the name
A SOSL search spans objects at once and its RETURNING clause keeps the response narrow; the record itself is then fetched by id when you know which one it was.
03
Cleaning up after an import that went in wrong
Query the affected records to see the damage, correct them as one bulk update, and bulk delete the rows that should never have been created. Each bulk call takes the whole list at once.
04
Reaching something the typed tools do not cover
List what objects exist, then go through the Tooling API for metadata work, Apex REST for custom endpoints, or the direct REST call for anything else the org exposes.
14 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
bulk_create_records
Creates multiple records of a specified SObject type in bulk.
bulk_delete_records
Deletes multiple records of a specified SObject type in bulk, given their IDs.
bulk_update_records
Updates multiple records of a specified SObject type in bulk. Each record must have an 'Id' field.
create_record
Creates a new record
delete_record
Deletes a record
get_object_fields
Retrieves field names and types for a Salesforce object. Use this to discover available fields before writing SOQL queries.
Output is CSV format: name,label,type,updateable
get_record
Retrieves a specific record by ID. Returns all fields - prefer SOQL with specific fields for token efficiency.
list_sobjects
Retrieves a list of all available Salesforce SObjects (standard and custom).
restful
Makes a direct REST API call to Salesforce
run_soql_query
Executes a SOQL query against Salesforce.
TOKEN OPTIMIZATION GUIDELINES:
- Always SELECT only the fields you need (never SELECT *)
- Use LIMIT to restrict results (start with LIMIT 10, increase if needed)
- Use WHERE clauses to filter data server-side
- Default output is CSV format (most token-efficient)
- Use format='json' only when you need nested data structure
- You can always run additional queries if you need more data
Example efficient query: SELECT Id, Name FROM Account WHERE IsActive = true LIMIT 20
run_sosl_search
Executes a SOSL search against Salesforce.
TOKEN OPTIMIZATION: Use RETURNING clause to limit fields and objects. Example: FIND {searchterm} RETURNING Account(Id, Name), Contact(Id, Name) LIMIT 10
tooling_execute
Executes a Tooling API request
update_record
Updates an existing record
BEFORE YOU INSTALL
What to know first
Tools you get
14 tools when we connected.
Requires
- uvx on PATH
- One Salesforce sign-in: SALESFORCE_ACCESS_TOKEN + SALESFORCE_INSTANCE_URL, SALESFORCE_CLIENT_ID + SALESFORCE_CLIENT_SECRET, or SALESFORCE_USERNAME + SALESFORCE_PASSWORD + SALESFORCE_SECURITY_TOKEN — or a Salesforce CLI default org
Where it runs
The documented client runs the server on your machine.
CONTEXT COST
Light Standing cost
Rated by the size of the 14 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
TERMINAL
uvx --from mcp-salesforce-connector salesforceMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"salesforce": {
"command": "uvx",
"args": [
"--from",
"mcp-salesforce-connector",
"salesforce"
]
}
}
}Set one sign-in method in the env block of this entry: SALESFORCE_ACCESS_TOKEN and SALESFORCE_INSTANCE_URL (OAuth, recommended); SALESFORCE_CLIENT_ID and SALESFORCE_CLIENT_SECRET (client credentials); or SALESFORCE_USERNAME, SALESFORCE_PASSWORD and SALESFORCE_SECURITY_TOKEN (legacy). With none of them set, the server uses your Salesforce CLI default org. Set SALESFORCE_DOMAIN to test to reach a sandbox.
