Skip to main content
Glama
shahin-raza

QueryBridge

by shahin-raza
OverviewSchemaSourceRelated ServersScoreDiscussionsAdmin
Python
Local

QueryBridge

QueryBridge is a local AI application that connects a MySQL database to an Ollama-powered language model through an MCP (Model Context Protocol) tool server. It lets you ask natural-language questions about your data while keeping database access structured, safe, and local.

What this project does

  • Exposes MySQL read operations through an MCP server

  • Uses a local Ollama LLM to generate a safe SQL query

  • Validates the SQL against the real table schema

  • Executes the query through the MCP tool layer

  • Passes the result back to the LLM for a human-readable answer

This app is intended for local development and experimentation with:

  • MySQL running in XAMPP

  • Ollama models running on the same machine

  • Python 3.12 + uv for project management

Related MCP server: MySQL MCP Server

Architecture

  • main.py – client entry point, asks a question and orchestrates the flow

  • mcp_server.py – MCP server that exposes safe database tools

  • db/connection.py – MySQL connection and SQL execution helpers

  • agent.py – Ollama LLM and prompt logic

  • .env – local environment variables for DB and model configuration

Prerequisites

Before running the application, make sure you have:

  • Python 3.12+

  • uv installed

  • Ollama installed and running locally

  • XAMPP or another local MySQL service running

  • A MySQL database named mcp_db

  • A table named employee_salary in that database

Install uv

If you do not have uv installed yet:

pip install uv

Or follow the official uv installation instructions for your OS.

Ollama setup

Make sure Ollama is installed and the service is running.

Pull the required models:

ollama pull llama3.1
ollama pull nomic-embed-text

Start Ollama:

ollama serve

If Ollama is already running in the background, you can skip this step.

The app uses these default values:

OLLAMA_BASE_URL=http://localhost:11434
LLM_MODEL=llama3.1
EMBEDDING_MODEL=nomic-embed-text

MySQL / XAMPP setup

This project is configured to work with a local MySQL instance, typically running under XAMPP.

Recommended DB values

Use the following values in your .env file:

DB_HOST=127.0.0.1
DB_NAME=mcp_db
DB_USER=root
DB_PASSWORD=root
DB_PORT=3306

Important notes:

  • Use 127.0.0.1 instead of localhost when running MySQL via XAMPP for better reliability on Windows

  • Ensure XAMPP MySQL is running before starting the app

  • The app expects the database mcp_db to exist

Example table

The application is designed around a table like:

CREATE TABLE employee_salary (
    emp_id INT PRIMARY KEY,
    emp_name VARCHAR(100),
    department VARCHAR(50),
    designation VARCHAR(50),
    base_salary DECIMAL(10,2),
    bonus DECIMAL(10,2),
    total_salary DECIMAL(10,2),
    joining_date DATE
);

Sample rows may look like:

INSERT INTO employee_salary (emp_id, emp_name, department, designation, base_salary, bonus, total_salary, joining_date)
VALUES
(1, 'Alice Johnson', 'Engineering', 'Software Developer', 75000.00, 5000.00, 80000.00, '2021-03-15'),
(2, 'Bob Smith', 'Engineering', 'DevOps Engineer', 72000.00, 4500.00, 76500.00, '2020-07-10');

Environment configuration

Create a .env file in the project root if it does not already exist.

Example:

DB_HOST=127.0.0.1
DB_NAME=mcp_db
DB_USER=root
DB_PASSWORD=root
DB_PORT=3306
OLLAMA_BASE_URL=http://localhost:11434
LLM_MODEL=llama3.1
EMBEDDING_MODEL=nomic-embed-text
TABLE_NAME=employee_salary

You can copy the values from .env.example if present in the project.

Running the project with uv

From the project folder:

cd C:\Users\srb3365\Documents\Projects\Python\querybridge

Install dependencies:

uv sync

Run the app:

uv run python main.py

When prompted, ask a question such as:

what is base salary of Bob Smith

The app will:

  1. read the schema from MySQL

  2. ask Ollama to generate a safe SELECT query

  3. call the MCP database tool

  4. return a natural-language answer based on the results

Quick validation commands

Check that MySQL is reachable:

uv run python -c "from db.connection import get_engine; from sqlalchemy import text; e=get_engine(); c=e.connect(); print(c.execute(text('SELECT 1 AS ok')).scalar_one()); c.close()"

Check the table exists:

uv run python -c "from db.connection import get_engine; from sqlalchemy import text; e=get_engine(); c=e.connect(); print(c.execute(text('SHOW TABLES')).fetchall()); c.close()"

Troubleshooting

MySQL connection refused

  • Confirm XAMPP MySQL service is running

  • Check that port 3306 is enabled

  • Use 127.0.0.1 instead of localhost

  • Validate the database credentials in .env

Ollama connection failed

  • Start Ollama with ollama serve

  • Pull the models: ollama pull llama3.1 and ollama pull nomic-embed-text

  • Confirm OLLAMA_BASE_URL points to the correct host and port

SQL generated is wrong

This app validates SQL against the current schema and rejects invalid table names or columns. If the model hallucinates a wrong table, the app will use a safe fallback for common employee salary questions.

Notes

This is a local-first project designed to demonstrate:

  • local model inference with Ollama

  • structured database access through MCP

  • MySQL read-only query execution

  • AI + tools pattern for business data questions

License

This project is provided for local learning and experimentation.

Available Tools

3 tools
get_employee_salary_schemaA

Return the schema of the employee_salary table for AI/agent use.

ParametersJSON Schema
NameRequiredDescriptionDefault

No parameters

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A3.7/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description indicates a read-only metadata retrieval operation ('Return the schema') and has no annotations to contradict this. However, it offers no additional behavioral context, such as whether the schema is cached, whether authentication is required, or any limitations. For such a simple tool, the basic behavior is adequately disclosed.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single concise sentence with the key action and resource front-loaded. The phrase 'for AI/agent use' is mildly redundant because the tool is already defined in an agent context, which prevents a perfect score.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a zero-parameter metadata tool with a provided output schema, the description sufficiently states the returned object. It lacks explicit guidance on when to call it relative to sibling tools, but the simplicity of the operation and the presence of an output schema make the definition essentially complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool has zero parameters, so the schema naturally covers all parameter semantics (100% coverage). Per the baseline for 0-parameter tools, the description does not need to add parameter details and none are missing.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb ('Return') and resource ('schema of the employee_salary table'), making the tool's function unambiguous. It clearly differs from sibling tools read_employee_salary and query_employee_salary, which would fetch data rather than metadata.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No guidance is given on when to use this tool versus the sibling read/query tools. The agent must infer from the tool name and context that this is the metadata lookup and the others are data access, which is not explicit.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

query_employee_salaryA

Run a read-only SQL query against the employee_salary table. Only SELECT statements are allowed.

ParametersJSON Schema
NameRequiredDescriptionDefault
sqlYes

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A3.9/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of disclosing behavior. It explicitly states the operation is read-only and limits input to SELECT statements, which is the most important safety behavior. It does not detail output formatting or query limits, but the available output schema helps cover return-value expectations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is two short sentences with no filler. It front-loads the core action and constraint immediately, and every sentence adds meaningful guidance.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a single-parameter tool with an output schema, the description is largely sufficient: it names the target table, restricts the query type, and clarifies read-only intent. It could be more complete by mentioning sibling-tool usage or query limits, but those are secondary for correct invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema provides only the parameter name 'sql' with no description, so schema coverage is 0%. The description compensates by clarifying that the sql parameter is a SQL query, must target the employee_salary table, and must be a SELECT statement. This gives the agent enough semantic grounding to construct the argument.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the verb and resource: run a read-only SQL query against the employee_salary table. It also specifies that only SELECT statements are allowed, which gives concrete scope. However, it does not explicitly differentiate itself from sibling tools like read_employee_salary or get_employee_salary_schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies this tool is for running ad-hoc read-only SQL queries against the employee_salary table. It provides a meaningful constraint ('Only SELECT statements are allowed') but gives no explicit guidance on when to choose this over read_employee_salary or get_employee_salary_schema.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

read_employee_salaryA

Read the latest rows from employee_salary. Keep the result small for LLM use.

ParametersJSON Schema
NameRequiredDescriptionDefault
limitNo

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A3.5/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the behavioral disclosure burden. 'Read' implies a non-mutating operation, and 'latest rows' suggests ordering by recency, but the description does not reveal limits on result size, pagination behavior, or any other side effects or constraints.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is two short sentences with no filler. The primary action is front-loaded, and the secondary sentence provides useful usage guidance without redundancy.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple read tool with one optional parameter and an output schema, the description is reasonably complete. However, it leaves the limit parameter semantics and the relationship to query_employee_salary implicit, so an agent must infer when this tool is the right choice.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters2/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, and the description does not explicitly explain the 'limit' parameter or how it relates to 'latest rows.' The phrase 'Keep the result small' indirectly implies limiting output, but it does not add clear parameter-level meaning beyond the schema's title and default.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Read') and the resource ('employee_salary'), and narrows the scope to 'latest rows,' which helps distinguish it from the more general query_employee_salary sibling. However, it does not explicitly name or contrast the sibling tools, so it stops short of full differentiation.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The instruction 'Keep the result small for LLM use' gives clear practical context for how the tool should be invoked. It does not explicitly state when to choose this tool over query_employee_salary or get_employee_salary_schema, so there is no exclusion guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. 3 tool updatesv0.1.0
    • First observedget_employee_salary_schema
    • First observedquery_employee_salary
    • First observedread_employee_salary
Install Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A4/5.0

Scored across 3 tools

Disambiguation5/5

Each tool has a clearly distinct purpose: schema introspection, small sample reads, and arbitrary read-only SQL queries. The read and query tools could theoretically overlap, but their descriptions make the intended use obvious.

Naming Consistency5/5

All tool names follow a consistent verb_noun pattern around the employee_salary entity. get, read, and query are different verbs but align naturally with each tool's specific action.

Tool Count5/5

Three tools is well-scoped for a focused server that exposes one table with schema, sample data, and ad-hoc querying. There are no unnecessary extras and nothing feels missing.

Completeness5/5

The server fully covers its intended read-only domain: understanding the schema, seeing recent rows, and running arbitrary SELECT queries. Write operations are not needed given the explicit read-only scope.

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

  • MySQL MCP Server

    caicongyang
    F
    license
    Not graded
    quality
    D
    maintenance
    A Model Context Protocol server that enables AI models to interact with MySQL databases through natural language, supporting SQL queries, table creation, and schema exploration.
    3
    -

  • MySQL MCP Server

    yuki777
    A
    license
    Not graded
    quality
    D
    maintenance
    A tool that connects to local MySQL databases and enables large language models (LLMs) to execute SQL queries through the Model Context Protocol (MCP).
    220
    1
    MIT

  • KatCoder MySQL MCP Server

    berthojoris
    F
    license
    Not graded
    quality
    D
    maintenance
    A secure MySQL Model Context Protocol server that enables AI agents to interact with MySQL databases through standardized operations. Features comprehensive security with SQL injection prevention, connection pooling, and configurable tool access for database operations.
    1
    -

  • MySQL MCP Server

    ImRieul
    A
    license
    A
    quality
    C
    maintenance
    A lightweight MySQL MCP server that enables LLMs to interact with databases through tools for schema inspection and query execution. It features LLM-friendly formatting, SSL support, and a secure read-only mode with query timeout protections.
    7
    53
    MIT

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/shahin-raza/querybridge'

If you have feedback or need assistance with the MCP directory API, please join our Discord server