Provides read-only SQL query execution against MySQL databases, supporting SELECT, SHOW, DESCRIBE, DESC, and EXPLAIN statements with parameterized queries, automatic row limits, and timeouts.
Provides read-only SQL query execution against PostgreSQL databases, supporting SELECT, SHOW, DESCRIBE, DESC, and EXPLAIN statements with parameterized queries, automatic row limits, and timeouts.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@db-query-mcp Query the analytics database for the average order value by month."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A Model Context Protocol (MCP) server for running read-only SQL queries across multiple MySQL and PostgreSQL databases. Built with TypeScript, runnable with Node.js, Bun, or pnpm.
Give an AI assistant (Claude Desktop, Copilot, Cursor, Zed, or any MCP-compatible client) safe, read-only access to your databases — without pasting credentials into the chat.
🗄️ MySQL & PostgreSQL — mix both engines in one config
🔒 Read-only by design — only SELECT/SHOW/DESCRIBE/DESC/EXPLAIN; DML/DDL keywords rejected
🔐 Credentials out of source — connection config lives in ~/.config/db-query-mcp/config.json, never in the repo
⏱️ Safety guardrails — automatic LIMIT, per-query timeout, parameterized queries
📦 Installable — npx, pnpm dlx, bunx, or global install; one binary, no source editing
🧩 Single config, many databases — each connection can expose multiple databases
Related MCP server: mysql-mcp
Node.js ≥ 20, Bun, or pnpm (any one)
npx db-query-mcp
# or
pnpm dlx db-query-mcp
# or
bunx db-query-mcpnpm install -g db-query-mcp
# or
pnpm add -g db-query-mcpThen run db-query-mcp directly.
git clone https://github.com/andychinghk01/db-query-mcp.git
cd db-query-mcp
pnpm install
pnpm build
node dist/index.jsdb-query-mcp reads connections from a JSON config file.
Path (first match wins):
DB_QUERY_MCP_CONFIG env var (absolute path)
~/.config/db-query-mcp/config.json
mkdir -p ~/.config/db-query-mcp
cp examples/config.example.json ~/.config/db-query-mcp/config.json
# edit with your real credentials
$EDITOR ~/.config/db-query-mcp/config.json{
"queryTimeoutMs": 30000, // optional, default 30000
"defaultRowLimit": 1000, // optional, default 1000
"connections": [
{
"name": "unique-connection-id", // shown in list_db
"description": "Human-readable note",
"type": "mysql", // "mysql" | "postgresql"; defaults to "mysql"
"host": "db.example.com",
"port": 3306,
"username": "readonly-user",
"password": "your-password",
"databases": [
{ "name": "actual_db_name", "description": "What this DB contains" }
]
}
]
}Field | Type | Required | Notes |
| array | yes | One entry per server you connect to |
|
| no | Defaults to |
| array | yes | Each becomes a queryable target in |
| number | no | Per-query timeout, default 30000 |
| number | no | Max rows returned, default 1000 |
💡 Use a dedicated read-only database user. The server enforces read-only SQL, but defense in depth matters — grant only
SELECT(andSHOW/EXPLAINwhere needed) to the user in this config.
Add the server to your MCP client config.
claude_desktop_config.json or .mcp.json){
"mcpServers": {
"db-query-mcp": {
"command": "npx",
"args": ["db-query-mcp"]
}
}
}{
"mcpServers": {
"db-query-mcp": {
"command": "bunx",
"args": ["db-query-mcp"]
}
}
}{
"mcpServers": {
"db-query-mcp": {
"command": "pnpm",
"args": ["dlx", "db-query-mcp"]
}
}
}{
"mcpServers": {
"db-query-mcp": {
"command": "npx",
"args": ["db-query-mcp"],
"env": { "DB_QUERY_MCP_CONFIG": "/custom/path/config.json" }
}
}
}list_dbLists every configured database with its connection name, engine type, host, port, and description. No arguments.
queryExecutes a read-only query.
Argument | Type | Required | Description |
| string | yes | Database name from |
| string | yes | Read-only SQL statement |
| array | no | Positional parameters — |
Statement allowlist — only SELECT, SHOW, DESCRIBE, DESC, EXPLAIN may start a query.
Keyword denylist — INSERT, UPDATE, DELETE, DROP, ALTER, CREATE, TRUNCATE, REPLACE, MERGE, CALL, EXEC, GRANT, REVOKE, LOCK, UNLOCK, LOAD DATA are rejected anywhere in the statement (after comment/quote normalization).
Row limit — SELECTs are wrapped as subselects and capped at defaultRowLimit (default 1000).
Timeout — queries abort after queryTimeoutMs (default 30s).
Parameterized queries — use ? (MySQL) or $1 (Postgres) placeholders; values never interpolated into SQL.
These guardrails reduce risk but are not a substitute for least-privilege database permissions. Always pair this server with a read-only DB user.
pnpm install
pnpm build # tsc -> dist/
pnpm dev # watch mode
pnpm start # node dist/index.js
pnpm start:bun # bun dist/index.jsMIT © Andy Ching
List configured MySQL and PostgreSQL databases with their names, types, and descriptions.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden. The verb 'List' implies a safe, read-only operation, but the description does not explicitly state side effects, permission requirements, or any operational constraints. It provides minimal behavioral context beyond the basic function, which is adequate but not comprehensive.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, focused sentence that immediately states the action and key output details. It contains no fluff or redundant information, making it highly concise and well-structured for quick agent reading.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool is very simple (no parameters, no output schema), and the description covers the core operation and output. It does not mention return value format or any limitations, but given the low complexity, the description is largely sufficient. Sibling differentiation is minimal, slightly reducing completeness.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, so the baseline is 4. The description need not explain parameter semantics, and the input schema (empty) confirms that no parameters exist. The description adds no unnecessary param details, which is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's function with a specific verb ('List') and resource ('configured MySQL and PostgreSQL databases'), and specifies the output fields (names, types, descriptions). This distinguishes it from the sibling tool 'query', which likely executes queries rather than enumerating databases.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage: use this tool to see available databases. However, it does not explicitly mention when not to use it or contrast with the sibling 'query' tool. There is no direct guidance on alternative tools, so it falls short of a clear 'when to use vs alternatives' explanation.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Execute a read-only SELECT query against a selected MySQL or PostgreSQL database. Supports optional parameterized values. Only SELECT/SHOW/DESCRIBE/EXPLAIN statements are allowed.
| Name | Required | Description | Default |
|---|---|---|---|
| sql | Yes | A read-only SQL SELECT statement. | |
| dbName | Yes | The name of the database to query (as shown by list_db). | |
| params | No | Optional positional parameters for the SQL statement. Use ? placeholders for MySQL or $1, $2, ... for PostgreSQL. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, but the description carries the burden by specifying the tool is read-only and enumerating allowed statement types (SELECT/SHOW/DESCRIBE/EXPLAIN). It also mentions parameterized values with placeholder syntax for MySQL/PostgreSQL, giving useful behavioral context. It doesn't describe output format, but for a query tool this is sufficient.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences, front-loaded with the core purpose, and no redundant information. Every sentence contributes meaning, making it an efficient description.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description covers the key aspects for a moderate-complexity tool: read-only constraint, allowed statements, and parameterization. It lacks an explicit mention of the result format, but given no output schema and the straightforward nature of a query tool, it is sufficiently complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, with parameter descriptions already documenting sql, dbName, and params. The description additionally reinforces the read-only constraint and clarifies that dbName is as shown by list_db, but doesn't add significant new semantics beyond the schema. Baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool executes read-only SQL queries (SELECT/SHOW/DESCRIBE/EXPLAIN) against a MySQL/PostgreSQL database, using a specific verb and resource. It distinguishes itself from sibling list_db, which lists databases, by focusing on query execution.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It clearly implies usage for read-only querying against a chosen database, referencing list_db for database selection. However, it doesn't explicitly state 'use this instead of list_db when you need to retrieve data,' making it a clear context but without explicit exclusionary guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v1.0.0list_dbqueryScored across 2 tools
The two tools have clearly distinct roles: list_db for discovery and query for executing SQL. There is no overlap in their purposes or outputs.
Both tools use a verb-first pattern, but list_db is verb_noun while query is just a verb. This is a minor deviation and still predictable.
With only 2 tools, the set feels thin, though each earns its place for the stated read-only query purpose. The count is at the lower boundary of typical MCP servers.
The tool surface fully covers the domain of read-only database querying: listing available databases and executing arbitrary SELECT/SHOW/DESCRIBE/EXPLAIN statements. No critical gaps exist.
MCP server for querying and analyzing data from ad platforms, analytics tools, and spreadsheets
Draxlr's remote MCP server connects AI assistants to your SQL databases and dashboards. Explore schemas, run read-only queries, manage saved queries and dashboards, and export results, all with row-level security so each user sees only their own data.
Query your org's data in natural language — read-only MCP access to SQL, NoSQL, files & warehouses.
