Salesforce Admin & Metadata MCP. Inspect schemas, audit permissions, run Apex code.

This server hooks your AI client into the guts of a Salesforce org. Forget jumping through Setup, Developer Console, or running reports manually; you just ask natural questions, and your agent handles deep governance tasks. It lets you inspect an entire org's structure without clicking through any menus.

Schema Inspection
Using sf_describe_object, the client pulls the full schema for any object name you give it. This detail includes every field’s data type, if it's required, and what relationship targets it has. You can check out picklist values or see exactly how a custom field is structured.

Data Model Mapping and Metadata Search
To map the whole system, your agent first runs sf_list_objects, which spits out every queryable and createable object—standard, custom, and managed components included. If you need to track down specific logic like flows or triggers, use sf_search_metadata to search organizational metadata by name or partial text match for Apex classes, flows, and other core system elements.

You can also narrow down team members using sf_search_users, which finds IDs and details based on their name, email, or username.

Security and User Auditing
When you're doing a security review, the agent pulls critical user data by running sf_list_users, giving you active users' profiles, roles, emails, and last login dates. To audit permissions more broadly, it retrieves a list of all security profiles using sf_list_profiles, which details what types of access control are configured for different user groups.

Capacity Management and Code Execution
Need to know if you're running low on resources? Run sf_org_limits to get current usage numbers against maximum limits for API calls, storage, SOQL queries, and other governance metrics. For testing or fixing data in one shot, the client executes anonymous Apex code using sf_execute_apex. This lets you run real code directly against the org's data set.