MCP Server - Oracle DB Context

Name: MCP Server - Oracle DB Context
Author: danielmeppiel

@danielmeppiel

代码仓库

关于 MCP Server - Oracle DB Context

MCP Server for working with large Oracle databases

#db-context #oracle-db #mcp-server

基本信息

分类

数据库

许可证

MIT

运行时

python

传输方式

stdio

发布者

danielmeppiel

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "db-context": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "ORACLE_CONNECTION_STRING",
        "-e",
        "TARGET_SCHEMA",
        "-e",
        "CACHE_DIR",
        "dmeppiel/mcp-db-context"
      ],
      "env": {
        "ORACLE_CONNECTION_STRING": "user/pass@localhost:1521/mydb",
        "TARGET_SCHEMA": "",
        "CACHE_DIR": ".cache"
      }
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is MCP Server - Oracle DB Context?

A Model Context Protocol (MCP) server that provides contextual database schema information for large Oracle databases. It enables AI assistants such as GitHub Copilot, Claude, and ChatGPT to understand and work with databases containing thousands of tables by intelligently caching and serving schema data.

How to use MCP Server - Oracle DB Context?

Install and configure it via Docker (recommended) or UV local installation in VSCode Insiders. Set environment variables like ORACLE_CONNECTION_STRING, TARGET_SCHEMA, CACHE_DIR, and optionally READ_ONLY_MODE and THICK_MODE. Start the server locally with uv run main.py. Once connected, AI assistants can call several MCP tools to interact with the database schema.

Key features of MCP Server - Oracle DB Context

Smart schema caching to minimize database queries
Targeted schema lookup for specific tables
Table search by name pattern matching
Foreign key relationship mapping between tables
Built specifically for Oracle databases
Read‑only mode enabled by default for security

Use cases of MCP Server - Oracle DB Context

AI assistant looks up schema for the EMPLOYEES table on demand
Developer searches for tables matching a pattern like “customer”
Analyst retrieves all foreign key relationships of the ORDERS table
Debugger gets source code of a PL/SQL stored procedure
Admin runs a read‑only SQL query safely through the AI assistant

FAQ from MCP Server - Oracle DB Context

What is the default security mode?

The server runs in read‑only mode by default (READ_ONLY_MODE=1), allowing only SELECT statements. Write operations (INSERT, UPDATE, DELETE) are blocked unless explicitly disabled by setting READ_ONLY_MODE="0".

What are the runtime requirements?

Python 3.12 or higher and access to an Oracle database. For thick mode, Oracle Instant Client is required. Docker usage includes all dependencies in the container.

How do I provide database credentials?

Set the ORACLE_CONNECTION_STRING environment variable in the format <db-username>/${input:db-password}@<host>:1521/<service-name>. The password is prompted securely via VSCode’s inputs configuration.

What tools are available for AI assistants?

Tools include get_table_schema, search_tables_schema, get_related_tables, run_sql_query, get_pl_sql_objects, get_object_source, get_table_constraints, get_table_indexes, rebuild_schema_cache, and more.

Can I run SQL queries through this server?

Yes, the run_sql_query tool lets you execute SQL queries. In default read‑only mode only SELECT is allowed; set READ_ONLY_MODE="0" to permit write operations.