MCP Server Use Cases

The MCP server supports controlled AI access to Autonomous AI Database by exposing tools defined using Select AI Agent and enforcing minimum-privilege policies. Explore some of the use cases.

Use cases include:

Use Cases	Description	Who/Persona	Task
Allow a chatbot to query metadata and call defined tools in Autonomous AI Database without opening SQL clients. Enable a multi-tenant environment where each database exposes a separate MCP endpoint with rate control.	Enable the MCP server for the target database. The server exposes one endpoint per database OCID and protects access with database security or credentials. This helps to allow trusted AI applications to reach approved database tools with security control.	Database Administrator ADMIN Database Engineer	Enable MCP Server
Register a Check Order Status tool for chatbots or customer support agents that runs specific SQL statements and provide order status. Provide a Get Customer Balance tool that runs a stored routine without exposing full financial tables.	Define a custom tool and bind it to an existing PL/SQL routine. This helps to provide a focused database action as a named tool so that AI applications use it with minimum privilege.	Privileged Database User	Register a custom tool
Integrate Autonomous AI Database with internal agents, such as Select AI Agent and OCI AI Agent that needs controlled access to metadata. Enable a developer’s AI application to query schema metadata in Autonomous AI Database during coding.	Configure the MCP client such as, OCI AI Agent or third-party clients such as, Claude Desktop, to reach the MCP server and sign in with database credentials. This enables chat or agent applications read schema information, object details, and query results.	Developer	Connect an AI client to the database
Analysts inspect table structures before writing queries. Automatically generate SQL queries without accessing sensitive data by using Select AI.	Use custom MCP tools to list schemas, list objects, get object details, and so on. This helps to understand table layouts and constraints without accessing sensitive data.	Data Analysts AI Agents	Run custom tools
Provide auditors with a clear record of actions taken by AI tools. Monitor MCP tools usage. Trace a suspicious access pattern to a specific user and tool.	Use logs and policies that capture user, tool, time, and scope. This helps to meet audit requirements and handle investigations.	System Compliance team	Audit MCP tool usage
Provide filtered access to sales data by region for field managers by using SQL custom agent tools. Allow finance teams to run queries on approved views with row-level masking by using custom agent tools.	Use the SQL tool in read-only mode with VPD policies. This helps to run data queries while seeing only permitted rows.	Database Administrators ADMIN Database Engineer	Secure data access using VPD policy.