AdoMcp

AdoMcp

AdoMcp is a Model Context Protocol (MCP) server that helps large language models (LLMs) understand database structure, read table comments, and execute SQL queries.

Category
Visit Server

README

AdoMcp

AdoMcp is a Model Context Protocol (MCP) server that helps large language models (LLMs) understand database structure, read table comments, and execute SQL queries.

AdoMcp 是一个基于 Model Context Protocol (MCP) 的数据库工具服务,帮助大型语言模型(LLM)理解数据库结构、读取表注释、执行 SQL 查询。

<!-- mcp-name: io.github.John0King/adomcp -->

MCP Tools

Tool Description
list_connections List configured database connections
add_connection Add (or replace) a database connection at runtime
remove_connection Remove a dynamically-added connection
list_objects List database objects (table/view/procedure/function/trigger/sequence/synonym, etc.)
get_table_schema Get table schema details (columns/types/nullability/PK/default/comments)
get_table_indexes Get table indexes
query_sql Execute read-only SQL and return CSV
execute_sql Execute write SQL (requires --allow-any-sql)

Recommended Tool Workflow (for LLM agents)

To reduce mistakes (wrong database/schema/object), use tools in this order:

  1. list_connections to discover available connections.
  2. If none are available, call add_connection.
  3. Before inspecting a table/view, call list_objects to locate schema + objectType + objectName.
  4. Use get_table_schema for column details (type, nullability, PK, default, comments).
  5. Use get_table_indexes when index/key design matters.
  6. Use query_sql only for read-only verification.
  7. Use execute_sql only when explicitly authorized and server is started with --allow-any-sql.

Oracle note: objects without owner prefix may be synonyms. Always confirm the real schema via list_objects first.

Supported Databases

Database Driver Comment support
SQL Server Microsoft.Data.SqlClient MS_Description extended properties
MySQL / MariaDB MySqlConnector TABLE_COMMENT / COLUMN_COMMENT
PostgreSQL Npgsql obj_description / col_description
SQLite Microsoft.Data.Sqlite — (SQLite has no native comments)
Oracle Oracle.ManagedDataAccess.Core ALL_TAB_COMMENTS / ALL_COL_COMMENTS (includes PUBLIC synonyms)

ORM support: Dapper · SqlSugarCore

Requirements


Quick Start

1. Configure database connections (optional)

Edit src/AdoMcp/appsettings.json and add pre-configured connections under the Databases array.
You can also skip this step entirely and let the LLM add connections dynamically via the add_connection tool.

"Databases": [
  {
    "Name": "mydb",
    "DbType": "SqlServer",
    "ConnectionString": "Server=localhost;Database=MyDb;User Id=sa;Password=***;TrustServerCertificate=true;",
    "Description": "Main business database"
  }
]

Supported DbType values: SqlServer | MySql | PostgreSql | Sqlite | Oracle

Security tip: Use .NET User Secrets or environment variables to manage connection strings in production.

2. Run the server

Automatic mode detection (recommended)

When stdin is redirected (i.e. launched by an MCP client), stdio mode is used automatically.
When run interactively in a terminal, HTTP/SSE mode is used automatically.

dotnet run --project src/AdoMcp

Specify mode manually

# stdio mode (all logs go to stderr; stdout carries only MCP JSON-RPC)
dotnet run --project src/AdoMcp -- --stdio

# HTTP/SSE mode (default: http://localhost:5100, MCP endpoint /mcp)
dotnet run --project src/AdoMcp -- --http

# Via environment variable
ADOMCP_MODE=http dotnet run --project src/AdoMcp

Enable execute_sql (write operations)

By default the execute_sql tool is disabled to prevent unauthorised writes.
Add --allow-any-sql to enable it:

dotnet run --project src/AdoMcp -- --allow-any-sql
# Combine with transport mode
dotnet run --project src/AdoMcp -- --http --allow-any-sql

3. Run via NuGet / dnx (.NET 10)

After the package is published to NuGet.org, you can run it without cloning the repo:

# Install as a global .NET tool once, then run directly
dotnet tool install -g AdoMcp
adomcp

# Or use dnx (.NET 10+) — installs and runs on demand
dnx AdoMcp
dnx AdoMcp -- --allow-any-sql

Dynamic connections at runtime (no config file needed)

LLMs can add new database connections during a session using add_connection:

User: Connect me to Oracle database oradb01
LLM → calls add_connection(
    connectionString = "Data Source=oradb01:1521/PROD;User Id=appuser;Password=***;",
    dbType = "Oracle",
    name = "prod-oracle",
    description = "Production Oracle DB"
)
→ returns: Connection 'prod-oracle' (Oracle) added successfully.
LLM → calls list_objects(connectionName = "prod-oracle")

Dynamically-added connections exist only for the lifetime of the process; restart the server or add the connection to appsettings.json for persistence.


Via dnx (after NuGet publish)

{
  "mcpServers": {
    "adomcp": {
      "command": "dnx",
      "args": ["-y","AdoMcp"]
    }
  }
}

HTTP mode

Start the server first:

dnx -y AdoMcp -- --http

Then configure the client:

{
  "mcpServers": {
    "adomcp": {
      "url": "http://localhost:5100/mcp"
    }
  }
}

Environment variables

All environment variables are prefixed with ADOMCP_ (override appsettings.json):

Variable Description
ADOMCP_MODE Transport mode: stdio or http (auto-detected when not set)
ADOMCP_URLS HTTP listen address, e.g. http://0.0.0.0:5100

MCP Registries

Official MCP Registry

This repository now includes server.json for the official MCP Registry with the server name io.github.John0King/adomcp.

To publish to the official MCP Registry:

  1. Publish the NuGet package AdoMcp to NuGet.org.
  2. Push a version tag such as v1.0.1, or manually run the Publish to NuGet and MCP Registry GitHub Actions workflow.
  3. The workflow publishes the NuGet package, authenticates with GitHub OIDC, and publishes server.json to the MCP Registry.

Build & Pack

# Build
dotnet build

# Pack as a NuGet tool (supports dnx)
dotnet pack src/AdoMcp -c Release -o ./nupkg

# Publish to NuGet.org (set NUGET_API_KEY first)
dotnet nuget push ./nupkg/AdoMcp.*.nupkg --source https://api.nuget.org/v3/index.json --api-key $NUGET_API_KEY

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured